SPSSX-L archives -- July 2006 (#205)

========================================================================= Date: Fri, 14 Jul 2006 09:11:30 -0400 Reply-To: Leslie Horst <lhorst@maguireassoc.com> Sender: "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU> From: Leslie Horst <lhorst@maguireassoc.com> Subject: Re: SPSSX-L Digest - 12 Jul 2006 to 13 Jul 2006 (#2006-192) List command output Content-Type: text/plain; charset="utf-8"

As a long-time SPSS user, I did a low-tech but effective solution, probably developed in a version long ago that didn't have "summarize."

String junk (a1). Compute junk = '/'. List var = v1 junk v2 junk v3 junk v4 junk.

Then when all of this goes into one column in Excel you can easily parse it using data/text to columns/delimited and then specifying the slash mark (or some other delimiter that you are sure does not exist in your data).

-----Original Message----- From: SPSSX(r) Discussion [mailto:SPSSX-L@LISTSERV.UGA.EDU] On Behalf Of Automatic digest processor Sent: Friday, July 14, 2006 12:02 AM To: Recipients of SPSSX-L digests Subject: SPSSX-L Digest - 12 Jul 2006 to 13 Jul 2006 (#2006-192)

There are 26 messages totalling 1173 lines in this issue.

Topics of the day:

1. export to xls 2. List Command output (3) 3. Likert scale (3) 4. SPSS and Terminal Services 5. Using Execute (2) 6. reliability 7. A Distinctly Non-Normal Distribution (3) 8. execute 9. SPSS-GLM query 10. percentile scores 11. About Mann-Whitney Test... 12. Logistic Regression queries 13. Working with dates - now pulled out ALL MY HAIR 14. Survival plot 15. Use of the weighted Kappa statistic 16. White paper for percentile rank formula (2) 17. Outreach Program Director Opening 18. Follow-up MANCOVA interaction?

----------------------------------------------------------------------

Date: Thu, 13 Jul 2006 09:30:01 +0300 From: vlad simion <vlad.simion@gmail.com> Subject: export to xls

Hi again,

I posted yesterday a message regarding an export to excel. No one has any sugesstions about how to export a string variable to an xls file via ODBC driver. My problem is that it only saves the first letter of the string and not the whole string.

Here is part of the export syntax:

save translate /connect= 'dsn=excel files;dbq=c:\documents and settings\ay08418\desktop\data\test.xls' /table=!quote(!n) /type=odbc /replace.

Thanks again!

Vlad.

-- Vlad Simion Data Analyst Tel: +40 720130611

------------------------------

Date: Thu, 13 Jul 2006 09:28:01 +0200 From: Mark Webb <targetlk@iafrica.com> Subject: List Command output

Hi all I'm using the List Command to list selected respondents and a few = variables. e.g. List Variables =3D Name Branch Dept Var1 Var2 Var3.

When I copy & past into Excel all variables listed go into one column. How can I get a SPSS output that will export into columns - like for = example the Frequencies command ? The print command is much the same.

Regards

Mark Webb

------------------------------

Date: Thu, 13 Jul 2006 11:44:55 +0300 From: Dominic Fernandes <dominic@spssme.com> Subject: Likert scale

Hi All,

I have a question. How do we analyze a Likert scale (consisting of 7 = responses) course evaluation questionnaire given to 25 teachers who = attended an in-service course? Shall we treat the responses as scale = variables and use parametric tests or shall we treat the responses as = ordinal variables and use the nonparametric test.

Thank you in advance for your assistance.

Dominic.

------------------------------

Date: Thu, 13 Jul 2006 11:55:27 +0100 From: Clare Gill <clare.gill@ucd.ie> Subject: SPSS and Terminal Services

Hello

We are considering a terminal services environment in UCD to deliver some of our applications to the UCD community. We would still be installing the windows version of SPSS so we would like to know if there has been or if there currently is any issues with delivering SPSS in this way.

As we have a large user community in UCD, we would also like to know what the recommended or expected hardware requirements would be to run SPSS in this environment as some of the analysis that can be done in SPSS can already use a lot of system resourses. We would consider 250/500 users as a guideline for system requirements.

-- Clare Gill Computing Services, University College Dublin, UCD Computer Centre, Belfield, Dublin 4. Tel: +353-1-716 2007 Fax: +353-1-2837077 http://www.ucd.ie/computing

------------------------------

Date: Thu, 13 Jul 2006 06:36:26 -0500 From: "Peck, Jon" <peck@spss.com> Subject: Re: Using Execute

The reason that the default behavior is to generate EXECUTE commands when working from the transformation dialogs is that without that, the Data Editor window does not update immediately, and new users found this behavior confusing. We wanted to make it easier to get started with SPSS.

Since you can turn this setting off, and since many users do not use syntax anyway, this is pretty harmless. The help for this option explains the efficiency issue, although many users never find their way to this spot.

Regards, Jon Peck

-----Original Message----- From: SPSSX(r) Discussion [mailto:SPSSX-L@LISTSERV.UGA.EDU] On Behalf Of Lisa Stickney Sent: Wednesday, July 12, 2006 9:50 PM To: SPSSX-L@LISTSERV.UGA.EDU Subject: Re: [SPSSX-L] Using Execute

Hi Richard,

> > Good ever-lovin' grief! Thank you, Lisa! I didn't even know that SPSS > inserted EXECUTEs following pasted syntax commands. (See below.) > > If SPSS puts EXECUTEs after every (pasted) transformation command, of > COURSE users will think they're necessary. > > Other readers: Did this happen to you, too? It's precisely the right > way to form precisely the wrong habit. >

I don't think it does it with every transformation command, but it definitely does it with COMPUTE, RECODE & COUNT. Plus it does it with some of the data commands -- FILTER, MATCH FILES, & ADD FILES that I know of.

> Beadle, ViAnn wrote at 12:53 PM 7/12/2006, > >>Although its not really obvious, you can turn off the EXECUTE command >>[that's pasted after pasted transformation commands] in Edit>Options >>under Transformation and Merge Operations. > > To be precise (since I had to look for it myself), it's > > Edit>Options>Data; then, under Transformation and Merge Operations, > + To get the EXECUTE pasted, select "Calculate values immediately"; > + To omit it, select "Calculate values before used". > (Which is the default?) >

Thanks to both ViAnn & Richard for pointing this out. I have happily changed my copy of SPSS.

As for the default, I believe it's "Calculate values immediately." The IT people have installed versions 11.5, 12, 13 & now 14 on my laptop, and they've all been set this way. So, unless it's just my installation or they're changing this upon installation (I doubt it), it's probably the default.

One other comment I have about this is option is that I think it would have little meaning to a newbie who's trying to learn SPSS sytax. Unless you're very familiar with SPSS, it's not clear how this relates to EXECUTE or why it might be important.

Best, Lisa

Lisa T. Stickney Ph.D. Candidate The Fox School of Business and Management Temple University Lts@temple.edu

------------------------------

Date: Thu, 13 Jul 2006 08:07:11 -0400 From: "Dates, Brian" <BDATES@swsol.org> Subject: Re: Likert scale

Dominic,

With seven (7) response categories, you can probably treat the data as interval in nature. I'm a purist and my principal area of work is in measurement development and standardization, so I generally follow the IRT approach and then do the analysis on the IRT scaled items. Likert actually developed a scaling method for his approach to items which would be worth your time for future endeavors. As part of your analysis, I would recommend at least performing a reliability analysis to check for internal consistency and the relationship of the items to total score, so you can identify any "clinker" items that might exist. Good luck.

Brian

Brian G. Dates, Director of Quality Assurance Southwest Counseling and Development Services 1700 Waterman Detroit, Michigan 48209 Telephone: 313.841.7442 FAX: 313.841.4470 email: bdates@swcds.org

> -----Original Message----- > From: Dominic Fernandes [SMTP:dominic@spssme.com] > Sent: Thursday, July 13, 2006 4:45 AM > To: SPSSX-L@LISTSERV.UGA.EDU > Subject: Likert scale > > Hi All, > > I have a question. How do we analyze a Likert scale (consisting of 7 > responses) course evaluation questionnaire given to 25 teachers who > attended an in-service course? Shall we treat the responses as scale > variables and use parametric tests or shall we treat the responses as > ordinal variables and use the nonparametric test. > > Thank you in advance for your assistance. > > Dominic. > > Confidentiality Notice for Email Transmissions: The information in this message is confidential and may be legally privileged. It is intended solely for the addressee. Access to this message by anyone else is unauthorised. If you are not the intended recipient, any disclosure, copying, or distribution of the message, or any action or omission taken by you in reliance on it, is prohibited and may be unlawful. Please immediately contact the sender if you have received this message in error. Thank you.

------------------------------

Date: Thu, 13 Jul 2006 07:59:57 -0500 From: "Beadle, ViAnn" <viann@spss.com> Subject: Re: List Command output

Use the summarize command which will give you a case listing in a table from which you can copy the cells:

summarize name branch dept var1 var2 var3 /cells none/format list nocasenum.

-----Original Message----- From: SPSSX(r) Discussion [mailto:SPSSX-L@LISTSERV.UGA.EDU] On Behalf Of Mark Webb Sent: Thursday, July 13, 2006 2:28 AM To: SPSSX-L@LISTSERV.UGA.EDU Subject: List Command output

Hi all I'm using the List Command to list selected respondents and a few variables. e.g. List Variables = Name Branch Dept Var1 Var2 Var3.

When I copy & past into Excel all variables listed go into one column. How can I get a SPSS output that will export into columns - like for example the Frequencies command ? The print command is much the same.

Regards

Mark Webb

------------------------------

Date: Thu, 13 Jul 2006 07:24:29 -0700 From: razina khayat <razinak@yahoo.com> Subject: reliability

Hi all, I'm trying to do a test-retest reliability analysis on attachment loss measurement. Attachment loss is measured in millimeters ranging from 0 to about 14. my question is how to do this within ± 1 mm accuracy. Also for validity assessment (using gold standard), how could we measure the same variable within 1 mm accuracy. thanks, razina

--------------------------------- Talk is cheap. Use Yahoo! Messenger to make PC-to-Phone calls. Great rates starting at 1¢/min.

------------------------------

Date: Thu, 13 Jul 2006 09:12:28 -0600 From: Stevan Nielsen <stevan_nielsen@byu.edu> Subject: A Distinctly Non-Normal Distribution

Dear Colleagues,

I have stumbled upon an interesting phenomenon: I have discovered that consumption of a valuable resource conforms to a very regular, reverse J-shaped distribution. The modal case in our large sample (N = 16,000) consumes one unit, the next most common case consumes two units, the next most common three units, the next most common four units -- and this is the median case, and so on. The average is at about 9.7 units, which falls between the 72nd and 73rd percentile in the distribution -- clearly NOT an indicator of central tendency.

I used SPSS Curve Estimation to examine five functional relationships between units consumed and proportion of consumers in the sample, testing proportion of consumers in the sample as linear, logarithmic, inverse, quadratic, or cubic functions of number of units consumed. I found that the reciprocal model, estimating proportion of cases as the inverse of units consumed, was clearly the best solution, yielding a remarkable, and very reliable R2 = .966. All five models were reliable, but the next best was the logarithmic solution, with R2 = .539; worst was the linear model, with R2 = .102.

These seems like a remarkably regular, quite predictable relationship. I've spent my career so enamored with normal distributions that I'm not sure what to make of this distribution. I have several questions for your consideration:

Do any of you have experience with such functions? (I believe it would be correct to call this a decay functions.)

Where are such functions most likely to occur in nature, commerce, epidemiology, genetics, healthcare, and so on?

What complications arise when attempting to form statistical inferences where such population distributions are present? (We have other measurements for subjects in this distributions, measurements which are quite nicely normal in their distributions.)

Your curious colleague,

lars nielsen

Stevan Lars Nielsen, Ph.D. Brigham Young University

801-422-3035; fax 801-422-0175

------------------------------

Date: Thu, 13 Jul 2006 08:31:18 -0700 From: "P. Scott Dixon" <psdixon@pdx.edu> Subject: Re: execute

I agree with ViAnn. With all respect to the SPSS wunderkinds, I think the use of executes is useful to find flaws in my code at the point they occur, and the cost in processor time is negligible with small to mid size samples. It is a strong tool in my, admittedly, less-than-perfect code.

Scott

-----------------------------

Date: Wed, 12 Jul 2006 07:04:19 -0500 From: "Beadle, ViAnn" <viann@spss.com> Subject: Re: reverse concat

OK, I'll take your bait.=20 =20 I don't actually use EXECUTE in syntax but when I'm work out gnarly = transformation problems I frequently Run Pending Transformations (which = generates your dreaded EXECUTE) from the data editor and see the effects = of my transformations immediately. My strategy is to work through the = problem a "executable" unit at a time and check out the unit by running = the pending transformations. Unless your dealing with wide (more than = 200 or so variables) or long (more than 10,000) or more cases the time = to actually do the transformation is much less than the think time to = step mentally through the process.=20 =20 What does that extra processing really cost you? It's not like the bad = old days when every job submittal at $1.00 a pop ate up my account at = the University of Chicago Comp Center.

________________________________

------------------------------

Date: Thu, 13 Jul 2006 16:55:13 +0100 From: Shweta Manchanda <shweta@manchanda.net> Subject: SPSS-GLM query

I would like to know what the estimated marginal means in the General Linear Model mean. I would be grateful to get any help.

Many thanks.

------------------------------

Date: Thu, 13 Jul 2006 09:57:03 +0300 From: alexwin@MAIL.UNIV.KIEV.UA Subject: Re: percentile scores

Small corrections: to obtain percentile scores using formulas described in psychometric literatuire (see, for instance, Crocker, L., & Algina, J. (1991). Introduction to classical and modern test theory. Orlando, FL: Harcourt Brace Jovanovich.), I would use slightly different options:

1) Under the Transform menu, choose Rank Cases… 2) Choose the variable containing the raw test scores 3) Click on the Ties… box and choose MEAN as the Rank Assigned to Ties 4) Click on the Rank Cases: Types… box and Check Fractional rank as % 5) Check Rankit under Proportion Estimation Formula 6) Uncheck Rank (Unless you want cumulative frequencies) 7) Click on OK

Dr. Alexander Vinogradov, Associate Professor Sociology and Psychology Faculty National Taras Shevchenko University Ukraine

>> Statisticsdoc <statisticsdoc@cox.net> wrote: Humphrey,

>> There are two ways to compute percentile scores in SPSS:

>> You can use menus to compute percentile ranks as follows:

>> 1.) Under the Transform menu, choose Rank Cases… >> 2.) Choose the variable containing the raw test scores >> 3.) Click on the Ties… box and choose High as the Rank Assigned to Ties >> 4.) Click on the Rank Cases: Types… box and Check Fractional rank as % >> 5.) Uncheck Rank (Unless you want cumulative frequencies) >> 6.) Click on OK >> 7.) SPSS will label the variable that contains the percentile ranks "PCT001".

------------------------------

Date: Thu, 13 Jul 2006 02:59:55 -0500 From: John McConnell <jmcconnell@APPLIED-INSIGHTS.CO.UK> Subject: Re: List Command output

Hi Mark

If you use the Case Summaries option (from memory this is in thenAnalyze>Reports menu and in syntax SUMMARIZE) you get a list inside a pivot table which is the key to Excel compatibility.

From there you should be able to cut/paste or even export the table to Excel to keep more of the formatting.

hth

john

John McConnell Applied Insights

---------- Original Message ---------------------------------- From: Mark Webb <targetlk@iafrica.com> Reply-To: Mark Webb <targetlk@iafrica.com> Date: Thu, 13 Jul 2006 09:28:01 +0200

>Hi all >I'm using the List Command to list selected respondents and a few variables. >e.g. List Variables = Name Branch Dept Var1 Var2 Var3. > >When I copy & past into Excel all variables listed go into one column. >How can I get a SPSS output that will export into columns - like for example the Frequencies command ? >The print command is much the same. > >Regards > >Mark Webb >

------------------------------

Date: Thu, 13 Jul 2006 06:55:35 -0500 From: "Oliver, Richard" <roliver@SPSS.COM> Subject: Re: Using Execute

The default behavior is for the GUI to generate an Execute statement = after transformation commands generated from dialogs (e.g. the Compute = dialog). This is the case for any command generated from the dialogs = that requires some subsequent command to read the data (which is why it = also happens with Match Files and Add Files). AFAIK, Filter doesn't = require a subsequent command to read the data, but the same dialog that = generates Filter syntax can also generate Select If syntax.=20 =20 There are reasons this was deemed to be the best default for the GUI, = but it may not be ideal for those attempting to learn syntax by pasting = dialog selections (although all those Executes will tell you which = commands require some other command to read the data).=20

________________________________

From: SPSSX(r) Discussion on behalf of Lisa Stickney Sent: Wed 7/12/2006 9:49 PM To: SPSSX-L@LISTSERV.UGA.EDU Subject: Re: Using Execute

Hi Richard,

I don't think it does it with every transformation command, but it definitely does it with COMPUTE, RECODE & COUNT. Plus it does it with = some of the data commands -- FILTER, MATCH FILES, & ADD FILES that I know of.

Thanks to both ViAnn & Richard for pointing this out. I have happily changed my copy of SPSS.

As for the default, I believe it's "Calculate values immediately." The = IT people have installed versions 11.5, 12, 13 & now 14 on my laptop, and they've all been set this way. So, unless it's just my installation or they're changing this upon installation (I doubt it), it's probably the default.

One other comment I have about this is option is that I think it would = have little meaning to a newbie who's trying to learn SPSS sytax. Unless = you're very familiar with SPSS, it's not clear how this relates to EXECUTE or = why it might be important.

Best, Lisa

Lisa T. Stickney Ph.D. Candidate The Fox School of Business and Management Temple University Lts@temple.edu

------------------------------

Date: Thu, 13 Jul 2006 18:13:33 +0200 From: Karl Koch <TheRanger@gmx.net> Subject: About Mann-Whitney Test...

Hi all,

The Mann-Whitney assumes independent samples. In Andy Fields book [2005] it also says that it assumes that datapoints are measured from different people - meaning that one person cannot partipate in both groups. I am not sure, if Andy Field here assumes that a datapoint always maps directory to a person (because of his background in Psychology). I find this a bit confusing and wouild like to ask you.

My datapoint are actually measures (on a 6-point scale) (which come from people but several from each person). Each person partipates in each of the two groups - from which I want to test stat. significance if they are different or not. The order with which people partipate in this two groups is counterballanced and should therefore be independent.

What do you think? Is this a case for Mann-Whitney then?

Best Regards, Karl --

Der GMX SmartSurfer hilft bis zu 70% Ihrer Onlinekosten zu sparen! Ideal für Modem und ISDN: http://www.gmx.net/de/go/smartsurfer

------------------------------

Date: Thu, 13 Jul 2006 11:31:32 -0500 From: Anthony Babinec <tbabinec@sbcglobal.net> Subject: Re: A Distinctly Non-Normal Distribution

Here are a couple general comments.

While the normal distribution might be a useful assumed distribution for errors in regression, there is no reason to think that it is necessarily useful for summarizing all phenomena out there in the world.

As you have described your data, they are counts. In other words, values are 1, 2, 3 etc., and not real values in some interval.

Are you looking at consumption in some fixed unit of time - say week, month, year? Given some assumptions, there are distributions such as the poisson that might be appropriate. It also could be the case that what you are studying represents a mixture of types, say usage types (low, medium, high), though that may or may not be the case here.

Pete Fader(Wharton) and Bruce Hardie(London Business School) have a nice course on probability models in marketing that is regularly given at AMA events.

-----Original Message----- From: SPSSX(r) Discussion [mailto:SPSSX-L@LISTSERV.UGA.EDU] On Behalf Of Stevan Nielsen Sent: Thursday, July 13, 2006 10:12 AM To: SPSSX-L@LISTSERV.UGA.EDU Subject: A Distinctly Non-Normal Distribution

Dear Colleagues,

Do any of you have experience with such functions? (I believe it would be correct to call this a decay functions.)

Where are such functions most likely to occur in nature, commerce, epidemiology, genetics, healthcare, and so on?

Your curious colleague,

lars nielsen

Stevan Lars Nielsen, Ph.D. Brigham Young University

801-422-3035; fax 801-422-0175

------------------------------

Date: Thu, 13 Jul 2006 11:25:50 -0400 From: K <statsq@HOTMAIL.CO.UK> Subject: Logistic Regression queries

Hi all,

I’m currently learning how to use Logistic Regression. I have a couple of queries that I can’t seem to find the answer to;

1) Most of my odds ratios are around 1, with some slightly higher at 15, and one at 386. This seems extremely high. Is there an explanation for such a high odds ratio? 2) I’ve read that high standard errors signal multcollinearity. Just how high is high though?

------------------------------

Date: Thu, 13 Jul 2006 14:04:11 -0300 From: Hector Maletta <hmaletta@fibertel.com.ar> Subject: Re: A Distinctly Non-Normal Distribution

The phenomenon you are describing seems to follow a Poisson distribution. There are also other asymmetrical distributions that apply to multiple phenomena in nature and society. The Poisson distribution was first tried by Bortkiewicz in the early 20th century to predict the number of Prussian soldiers killed annually by a horse-kick (or was it the number of times an average soldier would be kicked by a horse during his service time? I do not remember precisely), and applies to any event that could happen multiple times with decreasing probability of repetition. Another famous asymmetrical distribution, first rising to a maximum in some low value and then decreasing gradually towards higher values, is the Pareto equation for the distribution of income. He found few people with implausibly low incomes, then a maximum frequency around the most common wage level, and then a very long tail with decreasing frequencies as income increases all the way up to Bill Gates level. There is no reason to expect these phenomena to follow a symmetrical, let alone a normal Gaussian, distribution. Hector

-----Mensaje original----- De: SPSSX(r) Discussion [mailto:SPSSX-L@LISTSERV.UGA.EDU] En nombre de Stevan Nielsen Enviado el: Thursday, July 13, 2006 12:12 PM Para: SPSSX-L@LISTSERV.UGA.EDU Asunto: A Distinctly Non-Normal Distribution

Dear Colleagues,

Do any of you have experience with such functions? (I believe it would be correct to call this a decay functions.)

Where are such functions most likely to occur in nature, commerce, epidemiology, genetics, healthcare, and so on?

Your curious colleague,

lars nielsen

Stevan Lars Nielsen, Ph.D. Brigham Young University

801-422-3035; fax 801-422-0175

------------------------------

Date: Thu, 13 Jul 2006 13:13:42 -0400 From: "[Ela Bonbevan]" <bonbevan@AOL.COM> Subject: Working with dates - now pulled out ALL MY HAIR

Many thanks to all who helped with my date problems. I was able to use some of the syntax to fix some of my dates and I am most grateful for the help.

I must confess that I don't find the date import from Excel to SPSS very straightforward. The data in my excel database comes from two sources and the date formats are all over the place.

One set comes as 19950816 and this is the easiest. The other comes like this 1/20/1999 and displays like this 20-Jan-99. Today when I did the excel import it brought up the following 4 cells. Where 38058 was supposed to be 13 March 04 and 38205 was supposed to be 7 August 04. I have made numerous attempts to fix this by standardizing the formats in Excel before I bring the sheet in but formatting cells appears to have no effect.

08-JAN-04 38058 07.07.04 38205

Any help??

Diane

------------------------------

Date: Thu, 13 Jul 2006 14:00:19 -0400 From: Trinh Luong <trinhluong@rogers.com> Subject: Survival plot

Dear All, I'd like to change the y-axis labels on a Kaplan-Meier survival plot from proportions to percentages but I'm not sure how. Could someone help me with this problem? Many thanks, Trinh Luong, MSc. Erasmus MC Rotterdam

------------------------------

Date: Thu, 13 Jul 2006 15:06:52 -0300 From: "Della Mora, Marcelo" <marcelo.dellamora@citigroup.com> Subject: Likert scale

Hi Dominic

Try reliability analysis that tests internal consistency and relationship of each items to total score: correlation item-total, co-variance...

hope this helps you

Marcelo Della Mora

-----Original Message----- From: SPSSX(r) Discussion [mailto:SPSSX-L@LISTSERV.UGA.EDU]On Behalf Of Dominic Fernandes Sent: Thursday, July 13, 2006 5:45 AM To: SPSSX-L@LISTSERV.UGA.EDU Subject: Likert scale

Hi All,

I have a question. How do we analyze a Likert scale (consisting of 7 responses) course evaluation questionnaire given to 25 teachers who attended an in-service course? Shall we treat the responses as scale variables and use parametric tests or shall we treat the responses as ordinal variables and use the nonparametric test.

Thank you in advance for your assistance.

Dominic.

------------------------------

Date: Thu, 13 Jul 2006 19:34:01 +0100 From: Margaret MacDougall <margmacd@talk21.com> Subject: Use of the weighted Kappa statistic

Dear all

Apologies for cross-posting

Dear all

I plan to compare students' self-assessment scores (within the range A-E, or something similar) to that of the scores these students are allocated by their examiners (the 2nd examiners, say). I would like to consider using a weighted Kappa statistic. However, several questions have emerged about how best to proceed. I would be most grateful for assistance with each of them.

Could I please have some advice on how to decide between linear and quadratic weighting as the best choice in this case.

Further, whilst the raters fall conveniently into two perfectly distinguishable groups ('student and '2nd examiner'), these raters change from student to student, although occasionally, one 2nd examiner may rate more than one student. As I understand it, in its original form, the weighted Kappa statistic was designed not only under the assumption of there being two distinct classes of raters but also that these raters would not change from subject. I am therefore concerned that a standard weighted Kappa statistic may not be the correct one for me.

A related question is what formula to use for the standard error of the weighted Kappa which I should use.

To summarize, I have raised three main questions, the first of which relates to the type of weighting to assume, the second to the appropriateness of using a standard Kappa statistic for my problem and the third of which relates in turn to what formula to use for the standard error of the recommended Kappa statistic.

I look forward to receiving some much needed help.

Thank you so much

Best wishes

Margaret

--------------------------------- All new Yahoo! Mail "The new Interface is stunning in its simplicity and ease of use." - PC Magazine

------------------------------

Date: Thu, 13 Jul 2006 12:54:20 -0700 From: Frank Berry <frank.berry@yahoo.com> Subject: White paper for percentile rank formula

Hi,

Is there a white paper that explains a formula for each of the following percentiles in spss? When there is missing values in a numeric variable ranging from 0 to 100 (all integers), would the same formula for percentiles applies to missing values in that variable? For example, an equal percentile can be obtained for values of 49, 50 - missing and 51. Or in other cases, same value of 49 could have two or more percentiles.

TIA. Frank

/PERCENTILES= 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99

--------------------------------- Talk is cheap. Use Yahoo! Messenger to make PC-to-Phone calls. Great rates starting at 1¢/min.

------------------------------

Date: Thu, 13 Jul 2006 17:44:45 -0500 From: "Reutter, Alex" <areutter@spss.com> Subject: Re: White paper for percentile rank formula

See the algorithms document for the procedure(s) you're interested in.

http://support.spss.com/tech/Products/SPSS/Documentation/Statistics/algorithms/index.html

Using Guest/Guest as login/password.

Alex

> -----Original Message----- > From: SPSSX(r) Discussion [mailto:SPSSX-L@LISTSERV.UGA.EDU] On Behalf Of > Frank Berry > Sent: Thursday, July 13, 2006 2:54 PM > To: SPSSX-L@LISTSERV.UGA.EDU > Subject: White paper for percentile rank formula > > Hi, > > Is there a white paper that explains a formula for each of the following > percentiles in spss? When there is missing values in a numeric variable > ranging from 0 to 100 (all integers), would the same formula for > percentiles applies to missing values in that variable? For example, an > equal percentile can be obtained for values of 49, 50 - missing and 51. Or > in other cases, same value of 49 could have two or more percentiles. > > TIA. > Frank > > /PERCENTILES= 1 2 3 4 5 6 7 8 9 10 > 11 12 13 14 15 16 17 18 19 20 > 21 22 23 24 25 26 27 28 29 30 > 31 32 33 34 35 36 37 38 39 40 > 41 42 43 44 45 46 47 48 49 50 > 51 52 53 54 55 56 57 58 59 60 > 61 62 63 64 65 66 67 68 69 70 > 71 72 73 74 75 76 77 78 79 80 > 81 82 83 84 85 86 87 88 89 90 > 91 92 93 94 95 96 97 98 99 > >

------------------------------

Date: Thu, 13 Jul 2006 16:43:18 -0400 From: AMANDA THOMAS <athomas4@KENT.EDU> Subject: Outreach Program Director Opening

Hello, I have attached an Outreach Program Director job description. This position is located within the Bureau of Research Training and Services at Kent State University's College of Education, Health and Human Services. The department works primarily in evaluating grants and and assisting clients in understanding and using their data. Please pass this on to anyone you know who may be interested and/or qualified. Applications can be submitted online at https://jobs.kent.edu.

Amanda S. Thomas Bureau of Research Training and Services Kent State University 507 White Hall Kent, Ohio 44242 Phone: (330) 672-0788 Page 1

ADMINISTRATIVE/PROFESSIONAL JOB DESCRIPTION Developed for Equal Opportunity

CLASS TITLE: Outreach Program Director

KSU Class Code: AD47 EEO Code: 1C FLSA: Exempt Pay Grade: 08

BASIC FUNCTION:

To plan and direct all operational, administrative, and financial activities of a designated educational outreach program.

CHARACTERISTIC DUTIES AND RESPONSIBILITIES:

Develop and market designated educational outreach program.

Establish objectives; develop strategies for achieving objectives.

Oversee University-designated program budget.

Develop and/or revise program policies and procedures.

Serve as liaison to various constituent groups relative to program activities.

Develop linkages with external organizations, professional associations,and community groups to support program development.

Oversee the promotion and/or communication on all aspects of the program by strategically planning for the development and distribution of various materials.

Identify and secure program funding opportunities.

Evaluate and implement changes to programs.

Serve on various University committees.

Perform related duties as assigned.

Employees assigned in a sales unit and sales management role may be responsible for: Â· Identifying a list of prospects and managing the communication processes with the prospects among the unit staff members. Â· Meeting with client prospects (i.e. human resource managers,operations managers, decision-makers, executives, professionals and employees) to present information about Kent Stateâ€™s services and products. Â· Pricing solutions proposed, writing and reviewing proposals,contracts and training plans reflecting the organizationsâ€™ needs and Kent Stateâ€™s solutions. Â· Maintaining electronic records utilizing client relationship management software. Reporting on the client activity using the software. Â· Strategically planning for contracted sales activity, analyzing and reporting on prospect and client activity.

REPORTS TO:

Designated Supervisor

LEADERSHIP AND SUPERVISION:

Leadership of a small department, unit, or major function and/or direct supervision over administrative/professional employees.

MINIMUM QUALIFICATIONS:

Must Pass Criminal Background Check

Education and Experience:

Bachelorâ€™s degree in a relevant field; Masterâ€™s degree preferred; four to five years relevant experience.

Other Knowledge, Skills, and Abilities:

Knowledge of personal computer applications; budgeting; strategic planning.

Knowledge of business, employee development and/or organization development approaches and concepts.

Skill in written and interpersonal communication.

Demonstration of ability to maintain client relationships for retaining and/or cross-selling clients.

Demonstration of ability to strategically plan for contract sales services and for prospecting and client calls for a small unit

Ability to provide leadership and direction.

Evaluation experience. Experience writing grants, composing budget proposals for grant evaluations.

Advanced understanding of SPSS software.

PHYSICAL REQUIREMENTS:

Light work: Exerting up to 20 pounds of force occasionally, and/or up to 10 pounds of force frequently, and/or negligible amount of force constantly to move objects. Typically requires sitting, walking, standing, bending, keying, talking, hearing, seeing and repetitive motions.

Incumbent may be required to travel from building to building frequently and campus to campus occasionally. For those in a sales management role, incumbent will be required to travel to client sites frequently.

The intent of this description is to illustrate the types of duties and responsibilities that will be required of positions given this title and should not be interpreted to describe all the specific duties and responsibilities that may be required in any particular position. Directly related experience/education beyond the minimum stated may be substituted where appropriate at the discretion of the Appointing Authority. Kent State University reserves the right to revise or change job duties, job hours, and responsibilities.

FILE: AD47 SOURCE: Job Description ANALYST: ds DEPARTMENTAL AUTHORIZATION: A.Lane 8/7/05

------------------------------

Date: Thu, 13 Jul 2006 19:05:13 -0700 From: jamie canataro <nicole_can1204@yahoo.com> Subject: Follow-up MANCOVA interaction?

I have a question regarding a MANCOVA interaction and was wondering for some assistance.

I am performing a 2 X 3 MANCOVA and have received significant results on an interaction. I am looking to follow-up the significant interaction to determine which of the specific means are statistically different. Would this require me to create a syntax program? And if so, how can I acquire/create/borrow the syntax.

Thank you in advance.

Nicole Canning

--------------------------------- Yahoo! Music Unlimited - Access over 1 million songs.Try it free.

------------------------------

End of SPSSX-L Digest - 12 Jul 2006 to 13 Jul 2006 (#2006-192)

Back to: Top of message | Previous page | Main SPSSX-L page

listserv.uga.edu
Enterprise Information Technology Services
The University of Georgia
listhelp@uga.edu