change scores
Dear Colleagues, A student is evaluating a summer program for junior high students. One of the goals was to raise 'self esteem.' Measures were taken before the program, at the end, and a month later. He expected that the program would be most effective for those who entered with especially low self esteem. He divided the students into quartiles based on the pretest and compared these subgroups on change. He found that his hypothesis was supported - there was greatest positive change for those who entered in the lowest quartile. However, further examination showed a clear 'regression toward the mean' effect, including a small negative change for the group that entered in the highest quartile. (The test does not have great reliability.) Question: How should he analyze these data? I would appreciate some discussion of this situation. Thank you. Dale Berger = Instructions for joining and leaving this list and remarks about the problem of INAPPROPRIATE MESSAGES are available at http://jse.stat.ncsu.edu/ =
Re: eigenvalue: origin of term
A square matrix can be used to transform a vector in length, direction, or both. For example, a vector with two elements can be post-multiplied by a 2x2 matrix to generate a new two-element vector. If the vector is unchanged in direction, the vector is an 'eigen vector' for the matrix. Literally, this means "it's own vector" for the matrix. Such a special vector can be thought of as 'belonging to' the matrix. The ratio of new to old length is called an 'eigen value' for the matrix. A pxp matrix may have up to p eigen vectors and eigen values. Dale BergerProfessor and Dean, PsychologyClaremont Graduate University123 East Eighth StreetClaremont, CA 91711 FAX: 909-621-8905Phone: 909-621-8084http://www.cgu.edu/faculty/bergerd.html - Original Message - From: Karl L. Wuensch To: [EMAIL PROTECTED] Sent: Saturday, January 20, 2001 10:02 AM Subject: eigenvalue: origin of term Can any of you all enlighten me regarding the origin of the term "eigenvalue." Is it related to the German word "eigen?" +Karl L. Wuensch, Department of Psychology,East Carolina University, Greenville NC 27858-4353Voice: 252-328-4102 Fax: 252-328-6283[EMAIL PROTECTED] http://core.ecu.edu/psyc/wuenschk/klw.htm
Re: a brainteaser
Ooops, the formula doesn't work for small n and m. Consider n=2 and m=2. Each man will have at least one woman by his side no matter how you arrange them, so the expected number of men that will have at least one woman by his side is 2.00, and the probability of at least one woman by the side of a given man is 1.00. The formula says p = (2m)/(n+m-1) = (2x2)/(2+2-1) = 4/3 = 1.33! The expected number of women by the side of a given man is 1.33. Dale Berger - Original Message - From: Kenn Konstabel <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Sent: Tuesday, December 19, 2000 6:27 AM Subject: Re: a brainteaser > Elvis Rodriguez wrote: > > > Can anyone help with this? > > > > n men and m women are sitting around a ruond table. What is the > > Expected number of men that will have at least one woman by his side? > > provided that they sit randomly, the probability that a randomly selected > man has at least one woman by his side should be (2m)/(n+m-1) ? you can > multiply it by n to get the expected number you want. > > k > > > > = > Instructions for joining and leaving this list and remarks about > the problem of INAPPROPRIATE MESSAGES are available at > http://jse.stat.ncsu.edu/ > = > = Instructions for joining and leaving this list and remarks about the problem of INAPPROPRIATE MESSAGES are available at http://jse.stat.ncsu.edu/ =
Re: Significant interation effect for MANOVA
Hi Clark, You maybe are aware of this, but describing the 'location' of an interaction is not as simple as it might seem at first glance. For example, in a 2x2 ANOVA it is not possible to locate an interaction in any one cell, or even any two or three cells: the interaction involves a relationship among all four cells. You could eliminate any interaction by changing any one of the four cell means. The same reasoning applies to interaction in MANOVA - interaction involves at least four means and maybe more. Dale Berger - Original Message - From: Clark Dickin <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Sent: Friday, December 01, 2000 12:03 AM Subject: Significant interation effect for MANOVA > I think I am missing something really obvious here but I am stuck for an > answer. I have a significant main effect for both of my DV's and also > have a significant Interaction among the DV's but I am unable to > determine where the interaction comes from. More specifically, is there > a post hoc test that will allow me to determine the wherabouts of the > significant interaction, and if so where is it in SPSS Ver. 10? Or is it > enough to see that the order of the DV's changes for the two groups for > one of the measures (only one was shown to be significant). My output > shows that there is a significant interaction and that only one of the > DV's is significant, but I have three age groups and two experimental > groups. > > Any assiatance wouold be greatly appreciated > > > > = > Instructions for joining and leaving this list and remarks about > the problem of INAPPROPRIATE MESSAGES are available at > http://jse.stat.ncsu.edu/ > = > = Instructions for joining and leaving this list and remarks about the problem of INAPPROPRIATE MESSAGES are available at http://jse.stat.ncsu.edu/ =
Re: datasets w/equal R^2???
Colleagues: Here are the 4 pairs of X,Y variables from Anscombe's 1973 American Statistician paper. (The columns, in order, are X1, Y1, X2, Y2, etc. Calculate the means and SDs for each variable, and r for each pair. This is a nice example to emphasize the importance of plotting data before fitting a model.) Cheers, Dale Berger Psychology Department Claremont Graduate University 123 East Eighth Street Claremont, CA 91711 FAX: 909-621-8905 Phone: 909-621-8084 http://www.cgu.edu/faculty/bergerd.html 10.00 8.04 10.00 9.14 10.00 7.46 8.00 6.58 8.00 6.95 8.00 8.14 8.00 6.77 8.00 5.76 13.00 7.58 13.00 8.74 13.00 12.74 8.00 7.71 9.00 8.81 9.00 8.77 9.00 7.11 8.00 8.84 11.00 8.33 11.00 9.26 11.00 7.81 8.00 8.47 14.00 9.96 14.00 8.10 14.00 8.84 8.00 7.04 6.00 7.24 6.00 6.13 6.00 6.08 8.00 5.25 4.00 4.26 4.00 3.10 4.00 5.39 19.00 12.50 12.00 10.84 12.00 9.13 12.00 8.15 8.00 5.56 7.00 4.82 7.00 7.26 7.00 6.42 8.00 7.91 5.00 5.68 5.00 4.74 5.00 5.73 8.00 6.89 - Original Message - From: <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Sent: Wednesday, September 20, 2000 10:15 AM Subject: datasets w/equal R^2??? > There is an introductory example of two datasets with equal R^2 > (and possibly with equal coefficients?) but with > markedly different residuals. I can't for the life of me > remember the author's name that is associated with these > data, or where to find them. > > Any help would be appreciated. > > Bruce > > > Sent via Deja.com http://www.deja.com/ > Before you buy. > > > = > Instructions for joining and leaving this list and remarks about > the problem of INAPPROPRIATE MESSAGES are available at > http://jse.stat.ncsu.edu/ > = > = Instructions for joining and leaving this list and remarks about the problem of INAPPROPRIATE MESSAGES are available at http://jse.stat.ncsu.edu/ =
Re: On-line survey
- Original Message - From: Robert J. MacG. Dawson <[EMAIL PROTECTED]> To: Dale Berger <[EMAIL PROTECTED]> Cc: <[EMAIL PROTECTED]>; Art Kendall <[EMAIL PROTECTED]>; <[EMAIL PROTECTED]> Sent: Thursday, July 27, 2000 12:05 PM Subject: Re: On-line survey > > > Dale Berger wrote: > > > > Adding to Art's list: > > If one has email addresses for a population of interest and > > wishes to collect information that is not particularly > > sensitive, an internet format might work as well or better > > than a mail survey. Robert J. MacG. Dawson wrote: > Yes. However, those called "internet" rather than "email" surveys are > usually those that involve putting up a form on a web page and waiting > for a Heffalump to fall into it. IIRC, the original question dealt with > concerns such as whether other users could access the responses that > only make sense in such a context. The best that can be said for _these_ > is that they might work as well or better than the Shere Hite style > questionnaire- in- a- magazine survey. > Dale Berger replied: No one has suggested that wide open self-selection surveys allow generalization to a known population. Let's leave that poor dead horse to rest in peace. We need better vocabulary here. I recognize that my title "on-line survey" might imply a wide-open "y'all come" survey. In fact, I have in mind data collection using a web site where the address is given to a targeted population. There are clear advantages over a hard copy survey form sent through regular mail - greater convenience for the respondents, quicker responses, automatic data handling, etc. Perhaps "targeted on-line data collection" would be a better term. I don't think the term "email survey" is adequate, because email might not be involved except through automatic transmission of responses. The concern for privacy of responses is perhaps greater for a closed group of people who know each other than it would be for a wide open survey on the internet. DB > > There would still be problems of inference if the response rate was low > RD > Indeed. And "low" does not mean less than 10% of the sample, it means > less than (say) 90% of the _population_. The nonrespondents must be too > few to matter. This is a nonrandom sample, and the beginner's intuition > that a small sample cannot represent the whole is _correct_ for such. > > If the starting address list is truly randomized within the population, > things are a little better; in such a case, 90% or so of the address > list may be enough. It would also be enough if the address list were > chosen in a way that had no plausible connection to the question at > hand. > > Now, all this is verifiable. If the researcher using email (or snail > mail, or telephone interviews - pick your technology level) for a survey > can verify that the address list is randomly chosen from a well-defined > sampling frame, and that the nonresponse rate is low enough not to > affect the inference, the results may be usable. > > However, is this going to be the case? If in fact the address list is > chosen for convenience and may be significantly nonrandom, will the > study go ahead anyway? If the nonresponse rate is 75% - or even 50% or > 25% - will the study be dropped? If results are published, will they be > titled "Perceptions of Innumeracy Among American College Graduates" or > "Perceptions of Innumeracy Among Euphoric State University Alumni Who > Gave Their Email Addresses To The Alumni Office And Choose To Answer A > Certain Survey?" Only the latter would be accurate. > > The problem is very simple. Random sampling is a powerful technique > that allows us - despite the intuition of many intelligent people > without a statistical background - to make valid inferences from a small > sample to a larger population. > Nonrandom sampling is not; to do it, and to expect it to have the same > results as random sampling, is like building an airplane out of straw > and expecting it to fly. > > > -Robert Dawson > Dale Berger replied: Is a 90%+ response rate for a survey really necessary? One can argue that logically a 90% response rate leaves the possibility that the remaining 10% of the population would all have responded in a direction opposite to those who did respond. But we know that the real world does not work that way. Consider polls regarding elections, which produce verifiably accurate results with (typically) well under 90% response rates. (Polls may be adjusted to take into account known characteristics of non-respondents, but they still don't have the actual responses from those people - who COULD all vote for Ralp
Re: interaction effects
Hi Mike, The short answer to your question is that a higher order interaction tells you that lower order interactions and main effects may be misleading. That is, if you have a significant AxB interaction and a significant AxBxC interaction, the 3-way tells you that the size of the AxB interaction depends on the level of C. An AxB interaction may not even exist at some level of C, even though the overall AxB interaction is statistically significant. Hence, it would be misleading to report the AxB interaction as if it described your data well overall. If you have a 3-way interaction between AxBxC, there are three ways to interpret these results. They look quite different, but in fact are only different ways to say the same thing. The three interpretations are as follows: the BxC interaction depends on the level of A; the AxC interaction depends on the level of B; the AxB interaction depends on the level of C. Sometimes one way of looking at these interactions makes more sense than another. It may be helpful to plot the results in each of the three ways. Again, these are the same results, so you don't need to report all three interpretations. (If you do, be sure to acknowledge that you have an alternate way to look at the same data, rather than a new finding.) You may wish to follow up on the three-way interaction with 'simple-effects' tests, whereby you test the BxC interaction at each level of A separately, etc. The significant three-way interaction already established that these BxC interactions differ from each other, but the simple effects tests can help describe the size and direction of each of the simple 2-way BxC interactions at each level of A. Similarly, you can follow a 2-way BxC interaction at one level of A with simple effects tests whereby you test the effects of B at each level of C or vice versa. If the BxC interaction is significant, that tells you that the effects of B are not the same at each level of C, and vice versa. As before, one interpretation may make more sense than the other. Again, this gives you a chance to talk about the size and direction of effects. In general, statistical tests in ANOVA are most interpretable when you have only one df in the numerator for the F test. Abelson calls tests with more than one df in the numerator a 'blob' test, because the test does not tell you where the differences are. You can construct contrasts of interest to provide unambiguous tests. Another caution: The full 3-way ANOVA design may not be the best way to look at your data. It may be that the mean for one cell is quite different from all of the others. That could produce all sorts of significant tests in the ANOVA, including main effects, 2-way and 3-way interactions, which obscure the fundamental pattern in the data. Final advice: plot the data and verify the quality of your data and the appropriateness of your statistical model. Good luck, Dale Berger Professor and Dean, Psychology Claremont Graduate University 123 East Eighth Street Claremont, CA 91711 FAX: 909-621-8905 Phone: 909-621-8084 http://www.cgu.edu/faculty/bergerd.html - Original Message - From: Mike Hewitt <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Sent: Saturday, July 22, 2000 5:54 PM Subject: interaction effects > This is a multi-part message in MIME format. > --1EB6033EF40E264F6048C4E4 > Content-Type: text/plain; charset=us-ascii > Content-Transfer-Encoding: 7bit > > Members, > > I am looking for assistance in interpreting results of a study. It > involved the testing of three different music practicing conditions. I > performed a GLM-repeated measures with three factors (modeling, > self-listening, self-evaluation) in addition to a repeated measure > (test). There was a significant interaction for test x modeling x > self-evaluation. There was also a significant result for test x > modeling. Does the higher-order interaction negate the results the > "main effects" or lower-order interaction? > > Specifically, musicians who listened to "model" performance improved > their performance more than those that did not listen to a model. > Great. For the interaction (test x modeling x self-evaluation), the > modeling/self-evaluation group improved more than did the no > modeling/self-evaluation group (reinforcing the results for modeling > only). HOWEVER, the same result did not occur for the groups that did > not self-evaluate. They improved similarly to each other. > So...listening to a model is more effective than not listening to one > when there is no-self-evaluation. > > What then does this mean as far as the results for the test x modeling > result? I guess my question is does a higher-level interaction > "overrule" a lower-level interaction? > > I appreciate your help! > > TIA > > >
Re: On-line survey
Dear Colleagues, In response to private email, I'd like to clarify what I meant to include with the term 'on-line survey' and 'internet format.' By internet format I am thinking of a web site where someone can respond by pointing and clicking, and the data are stored in a file for import into a data analysis program. email is not an essential part of this - we could notify potential respondents of the web site by other methods of communication. 'on-line survey' means that the survey is completed on-line. It might be better to call this web-page data collection, and separately specify procedures to determine respondents, methods from highly controlled to y'all come. We could use some new vocabulary here to make important distinctions. I was a bit taken aback by some very strong negative responses to my original question. It seems that some folks feel they are surrounded by idiots, and interpret ambiguities accordingly. Oh, well. It does make the edstat list more exciting. Dale Dale Berger wrote: > Adding to Art's list: > > If one has email addresses for a population of interest and wishes to > collect information that is not particularly sensitive, an internet format > might work as well or better than a mail survey. There would still be > problems of inference if the response rate was low, and perhaps the filters > are different than a mail survey. (There is an empirical question waiting > to be explored.) > > (somewhat shakier ground...) If one wishes to learn more about a rare > population (left-handed Lithuanians living in Chicago?), the internet could > be used to find multiple cases (or at least respondents who say they meet > the criteria). Granted, one would not be able to generalize to a > population. However, with careful screening, the responses could be treated > as case studies. If we had direct access to only a very few cases, these > supplementary cases might give us useful ideas for further research. > > It may be possible to open up direct contact with interesting cases > (identical twins raised apart from birth?). > > Dale Berger > > - Original Message - > From: Art Kendall <[EMAIL PROTECTED]> > To: <[EMAIL PROTECTED]> > Sent: Monday, July 17, 2000 12:29 PM > Subject: Re: On-line survey > > > There are a _few_ circumstances in which valid information can be > gathered > > using a survey form on the internet. I can think of three. Perhaps others > can > > identify more such circumstances. > > > > One circumstance is to have volunteers on the internet respond to a > survey form > > as one part of the pretesting of a survey instrument. For example, you can > see > > if there are a lot of "other" responses with a particular write-in answer. > This > > would allow you change the response categories for that question. You > might also > > find different ways the respondents managed to misunderstand questions. > > > > The second circumstance is when the population of interest is on a defined > part > > of the internet (perhaps an intranet), for example, employees of a company > or > > agency. > > > > The third circumstance is to gather some data to practice data cleaning > and > > preparation skills. > > > > Dale Berger wrote: > > > > > Dear Colleagues, > > > > > > I have a student who would like to collect data using a survey form on > the > > > internet. He would like to have data collected in a format that can be > > > imported easily into SPSS. Other desired features: confidentiality, > though > > > not anonymity - that is, he would like to assure users that other users > will > > > not have access to individual responses. He also would like to assure > that > > > no one responds more than one time, and ideally be able to track who has > > > responded. > > > > > > I know there are commercial firms that will do this work, but I would > prefer > > > to have the information collected on our computer, and to be close to > the > > > data collection process. What do we need to know to do this ourselves? > Has > > > anyone had experience with software to make this easier? > > > > > > Thanks for any advice. Dale > > > > > > Dale Berger > > > Professor and Dean, Psychology > > > Claremont Graduate University > > > 123 East Eighth Street > > > Claremont, CA 91711 > > > > > > FAX: 909-621-8905 > > > Phone: 909-621-8084 > > > http://www.cgu.edu/faculty/bergerd.html > > > > > > ===
Re: On-line survey
Adding to Art's list: If one has email addresses for a population of interest and wishes to collect information that is not particularly sensitive, an internet format might work as well or better than a mail survey. There would still be problems of inference if the response rate was low, and perhaps the filters are different than a mail survey. (There is an empirical question waiting to be explored.) (somewhat shakier ground...) If one wishes to learn more about a rare population (left-handed Lithuanians living in Chicago?), the internet could be used to find multiple cases (or at least respondents who say they meet the criteria). Granted, one would not be able to generalize to a population. However, with careful screening, the responses could be treated as case studies. If we had direct access to only a very few cases, these supplementary cases might give us useful ideas for further research. It may be possible to open up direct contact with interesting cases (identical twins raised apart from birth?). Dale Berger - Original Message - From: Art Kendall <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Sent: Monday, July 17, 2000 12:29 PM Subject: Re: On-line survey > There are a _few_ circumstances in which valid information can be gathered > using a survey form on the internet. I can think of three. Perhaps others can > identify more such circumstances. > > One circumstance is to have volunteers on the internet respond to a survey form > as one part of the pretesting of a survey instrument. For example, you can see > if there are a lot of "other" responses with a particular write-in answer. This > would allow you change the response categories for that question. You might also > find different ways the respondents managed to misunderstand questions. > > The second circumstance is when the population of interest is on a defined part > of the internet (perhaps an intranet), for example, employees of a company or > agency. > > The third circumstance is to gather some data to practice data cleaning and > preparation skills. > > Dale Berger wrote: > > > Dear Colleagues, > > > > I have a student who would like to collect data using a survey form on the > > internet. He would like to have data collected in a format that can be > > imported easily into SPSS. Other desired features: confidentiality, though > > not anonymity - that is, he would like to assure users that other users will > > not have access to individual responses. He also would like to assure that > > no one responds more than one time, and ideally be able to track who has > > responded. > > > > I know there are commercial firms that will do this work, but I would prefer > > to have the information collected on our computer, and to be close to the > > data collection process. What do we need to know to do this ourselves? Has > > anyone had experience with software to make this easier? > > > > Thanks for any advice. Dale > > > > Dale Berger > > Professor and Dean, Psychology > > Claremont Graduate University > > 123 East Eighth Street > > Claremont, CA 91711 > > > > FAX: 909-621-8905 > > Phone: 909-621-8084 > > http://www.cgu.edu/faculty/bergerd.html > > > > = > > Instructions for joining and leaving this list and remarks about > > the problem of INAPPROPRIATE MESSAGES are available at > > http://jse.stat.ncsu.edu/ > > = > > > > = > Instructions for joining and leaving this list and remarks about > the problem of INAPPROPRIATE MESSAGES are available at > http://jse.stat.ncsu.edu/ > = > = Instructions for joining and leaving this list and remarks about the problem of INAPPROPRIATE MESSAGES are available at http://jse.stat.ncsu.edu/ =
Re: On-line survey
Dear Colleagues, Thank you for many useful comments and suggestions regarding on-line data collection. Several people expressed concerns about the quality of data collected on-line, even to the point of categorically denouncing the technique. On-line surveys certainly have serious limitations, though there may be some applications where on-line data collection is superior to other methods. Consider a survey of recent alumni from a graduate program, where email addresses are available for all of the potential respondents. I seems to me that an on-line survey of this population could be better than a mail survey. The response rate may be greater, and with electronic transcription of data, there may be fewer coding errors. An on-line survey is certainly faster and less expensive. In this application there may be little risk of duplicate responses, and not much concern over confidentiality. In my view, categorical dismissal of on-line data collection is not warranted, though it will be important to assess threats to validity, find ways to counter them, and consider them in the analysis and interpretation of data. Dale Berger Professor and Dean, Psychology Claremont Graduate University 123 East Eighth Street Claremont, CA 91711 FAX: 909-621-8905 Phone: 909-621-8084 http://www.cgu.edu/faculty/bergerd.html Original message (in part): http://jse.stat.ncsu.edu/ =
On-line survey
Dear Colleagues, I have a student who would like to collect data using a survey form on the internet. He would like to have data collected in a format that can be imported easily into SPSS. Other desired features: confidentiality, though not anonymity - that is, he would like to assure users that other users will not have access to individual responses. He also would like to assure that no one responds more than one time, and ideally be able to track who has responded. I know there are commercial firms that will do this work, but I would prefer to have the information collected on our computer, and to be close to the data collection process. What do we need to know to do this ourselves? Has anyone had experience with software to make this easier? Thanks for any advice. Dale Dale Berger Professor and Dean, Psychology Claremont Graduate University 123 East Eighth Street Claremont, CA 91711 FAX: 909-621-8905 Phone: 909-621-8084 http://www.cgu.edu/faculty/bergerd.html = Instructions for joining and leaving this list and remarks about the problem of INAPPROPRIATE MESSAGES are available at http://jse.stat.ncsu.edu/ =
Re: Rates and proportions
Yet, p=0 is a special case where an outcome is impossible. A reasonable confidence interval for p should not include zero if the outcome has been observed in a sample. Not so? -Dale - Original Message - From: Donald Burrill <[EMAIL PROTECTED]> To: Dale Berger <[EMAIL PROTECTED]> Cc: <[EMAIL PROTECTED]>; <[EMAIL PROTECTED]> Sent: Wednesday, June 21, 2000 12:27 AM Subject: Re: Rates and proportions > On Tue, 20 Jun 2000, Dale Berger wrote: > > > If we observe one escape out of 1250 inmates, why can't we reliably > > rule out zero as the population escape rate? > > Because k = 1 (for n = 1250) is not significantly different from k = 0. > > > The normal approximation to the binomial may not be appropriate here. > > No, I don't expect it is. So use the binomial distribution. > > That's supposing that one wants a statistical argument. If a purely > logical argument suffices, it is indeed the case that a counterexample > demonstrates the falsity of a proposition. But it may still be not > unreasonable to ask, with what probability may one observe one (or more) > escapes outof n=1250 (or whatever n actually applies), if the true > probability of an escape is ? > (I specify non-zero only because it's difficult to carry out some > computations when p=0 exactly.) > And it is certainly reasonable to ask what confidence interval on p is > associated with k = 1. > -- Don. > > Donald F. Burrill [EMAIL PROTECTED] > 348 Hyde Hall, Plymouth State College, [EMAIL PROTECTED] > MSC #29, Plymouth, NH 03264 603-535-2597 > 184 Nashua Road, Bedford, NH 03110 603-471-7128 > > > > === > This list is open to everyone. Occasionally, less thoughtful > people send inappropriate messages. Please DO NOT COMPLAIN TO > THE POSTMASTER about these messages because the postmaster has no > way of controlling them, and excessive complaints will result in > termination of the list. > > For information about this list, including information about the > problem of inappropriate messages and information about how to > unsubscribe, please see the web page at > http://jse.stat.ncsu.edu/ > === > === This list is open to everyone. Occasionally, less thoughtful people send inappropriate messages. Please DO NOT COMPLAIN TO THE POSTMASTER about these messages because the postmaster has no way of controlling them, and excessive complaints will result in termination of the list. For information about this list, including information about the problem of inappropriate messages and information about how to unsubscribe, please see the web page at http://jse.stat.ncsu.edu/ ===
Re: Rates and proportions
Hi Don et al., If we observe one escape out of 1250 inmates, why can't we reliably rule out zero as the population escape rate? The normal approximation to the binomial may not be appropriate here. Dale Berger Professor and Dean, Psychology Claremont Graduate University 123 East Eighth Street Claremont, CA 91711 FAX: 909-621-8905 Phone: 909-621-8084 http://www.cgu.edu/faculty/bergerd.html - Original Message - From: Donald Burrill <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Cc: <[EMAIL PROTECTED]> Sent: Tuesday, June 20, 2000 2:49 PM Subject: Re: Rates and proportions > On Tue, 20 Jun 2000 [EMAIL PROTECTED] wrote: > > > Hello, I "inherited" the reporting system for our escapes and have some > > questions about how data has been reported in the past. ; ; > "Unreliable" or "useless"? Well, the basic graininess in a rate > is one escapee more (or less) than was reported. A rate of .08 per 100 > is about 1 out of 1250. If the data on which the rate was based were 1 > escapee out of 1250 inmates, one cannot _reliably_ tell the rate from > zero. If the data were 13 escapees out of 16,200 inmates, one would have > more faith in the rate, at least insofar as representing a small value > different from (not equal to!) zero. Unfortunately, the rate itself > does not tell one how grainy the data were. > === This list is open to everyone. Occasionally, less thoughtful people send inappropriate messages. Please DO NOT COMPLAIN TO THE POSTMASTER about these messages because the postmaster has no way of controlling them, and excessive complaints will result in termination of the list. For information about this list, including information about the problem of inappropriate messages and information about how to unsubscribe, please see the web page at http://jse.stat.ncsu.edu/ ===
Re: Sample size: way tooo big?
When we focus on estimates of effect sizes and the stability of those estimates, we are delighted to have a huge sample. Don't focus on statistical significance. === This list is open to everyone. Occasionally, less thoughtful people send inappropriate messages. Please DO NOT COMPLAIN TO THE POSTMASTER about these messages because the postmaster has no way of controlling them, and excessive complaints will result in termination of the list. For information about this list, including information about the problem of inappropriate messages and information about how to unsubscribe, please see the web page at http://jse.stat.ncsu.edu/ ===
Biostatistician: Western University of Health Sciences, Pomona, CA
Biostatistician and/or Epidemiologist Western University of Health Sciences, a graduate university of medical sciences and osteopathic medicine, seeks a full time biostatistician and/or epidemiologist who serve as a University-wide resource for faculty research project design and evaluation, and teach interdisciplinary research methods courses. Responsibilities: To integrate the effective use of clinical research methods across the University; to serve as a consultant for faculty and other members of the college in the design and implementation of subject matter research; to serve as a consultant for the office of strategic planning and other members of the university in the design and implementation of educational outcomes research; to teach interdisciplinary research methods course(s). Qualifications: Master's degree and two years teaching and research experience. Sound conceptual understanding of data analysis (categorical, longitudinal), clinical trials (outcomes research, evidence-based approach), and modeling. Ability to identify and study factors leading to diseases and disability, with the ultimate goal of prevention and control. Ability to function independently and collaboratively on research projects and with other scientists, and good verbal and written communication skills. Experience with experimentation in education and the health sciences, and working knowledge of statistical software (SPSS, SAS) and standard PC software (Excel, Access, Word). Preference: A Master's degree in Public Health or near completion of a doctoral degree in epidemiology, statistics, biostatistics, or related field; Ph.D. preferred. Review of applications begins March 20, 2000. To learn about Western University of Health Sciences, please visit the Web site at www.westernu.edu. Please send resume and cover letter to: Dr. Carl Trinca, Executive Vice President Western University of Health Sciences 309 East Second Street/College Plaza Pomona, California 91766-1854 === This list is open to everyone. Occasionally, less thoughtful people send inappropriate messages. Please DO NOT COMPLAIN TO THE POSTMASTER about these messages because the postmaster has no way of controlling them, and excessive complaints will result in termination of the list. For information about this list, including information about the problem of inappropriate messages and information about how to unsubscribe, please see the web page at http://jse.stat.ncsu.edu/ ===
Re: SPSS lab guide for mulitvariate stats.
Consider Tabachnick and Fidell (Using Multivariate Statistics). Not a lab guide, but it does include data sets and exercises, sample input and printouts, and even sample write-ups of results. Dale Berger - Original Message - From: Paul W. Jeffries <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]>; <[EMAIL PROTECTED]>; <[EMAIL PROTECTED]> Sent: Friday, February 11, 2000 11:29 AM Subject: SPSS lab guide for mulitvariate stats. > I am looking for a lab guide to teach SPSS mulitivariate stats. This book > is for psychology graduate students. Ideally, the book would have data > sets and exercises that cover multiple regression and correlation, > factorial analysis, cluster analysis, and discriminant analysis. Can > anyone recommend a book? > > Apologies for cross posting. > > Paul W. Jeffries > Department of Psychology > SUNY--Stony Brook > Stony Brook NY 11794-2500 > > > > > > === > This list is open to everyone. Occasionally, people lacking respect > for other members of the list send messages that are inappropriate > or unrelated to the list's discussion topics. Please just delete the > offensive email. > > For information concerning the list, please see the following web page: > http://jse.stat.ncsu.edu/ > === > === This list is open to everyone. Occasionally, people lacking respect for other members of the list send messages that are inappropriate or unrelated to the list's discussion topics. Please just delete the offensive email. For information concerning the list, please see the following web page: http://jse.stat.ncsu.edu/ ===
Re: Hotdecking methods
Hi Junjia, An example of hot-deck imputation: Suppose that the design of a random survey requires responses from four specific households in a particular neighborhood, but no one can be located in Household #4. The responses from Household #3 are duplicated and used as the imputed values for Household #3. The logic of this 'hot deck' procedure is that values from a neighbor are likely to be better estimates than overall sample means. cheers, Dale Berger Dale Berger Professor and Dean, Psychology Claremont Graduate University 123 East Eighth Street Claremont, CA 91711 FAX: 909-621-8905 Phone: 909-621-8084 Statistics web site: www.wise.cgu.edu http://www.cgu.edu/faculty/bergerd.html - Original Message - From: Junjia Liu <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Sent: Sunday, February 06, 2000 12:05 PM Subject: Hotdecking methods > Could someone tell me what hotdecking methods are in imputation techniques. > Thanks. > > Junjia > [EMAIL PROTECTED] > === This list is open to everyone. Occasionally, people lacking respect for other members of the list send messages that are inappropriate or unrelated to the list's discussion topics. Please just delete the offensive email. For information concerning the list, please see the following web page: http://jse.stat.ncsu.edu/ ===
Re: outside of class work
The problem in a nutshell - (1) practice with statistical applications outside of class, as with homework exercises, is essential; (2) we can't be sure who did work that was done outside of class; (3) If we don't grade homework, many students won't do it, despite good intentions. There may be no perfect solution. My approach is to allow, and even encourage, students to work together on homework in teams of two or three. Include issues for discussion. Promise that exams will include some questions based on homework problems and discussion issues. Grade homework for feedback, but don't give it much weight. Collaboration on homework can work very well. As we all know, teaching is a great way to solidify one's own learning. Despite the appeal of take-home exams, I no longer use them. I find them difficult to write and grade, there is the possibility of collaboration at some level, and some students put in tremendous amounts of time that distorts their lives and goes way beyond what is possible for many other students. That doesn't seem fair to me. Holiday Greetings to all! Dale Berger Professor and Chair, Psychology Claremont Graduate University 123 East Eighth Street Claremont, CA 91711 FAX: 909-621-8905 Phone: 909-621-8084 Statistics web site: http://www.wise.cgu.edu - Original Message - From: dennis roberts <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]>; <[EMAIL PROTECTED]> Sent: Sunday, December 26, 1999 5:47 PM Subject: outside of class work > At 10:29 PM 12/26/99 +, T.-S. Lim wrote: > > >I don't like take-home exams at all. It's very hard (almost impossible) to > >ensure that each student does the exam alone. It's true even for graduate > >students (I'm speaking from my own personal experiences). > > whether it is an exam ... or some other kind of work (projects, papers, > etc.) ... one really never knows IF it is his/her own work ... UNLESS one > takes the time to probe and ask followup questions about it ... but nobody > does ... > > however, you HAVE to give outside of class work (whether you figure it into > their grade or not) since there is no other time to be able to create > ... organize ... think for an extended period of time ... > == > dennis roberts, penn state university > educational psychology, 8148632401 > http://roberts.ed.psu.edu/users/droberts/droberts.htm >
Re: teaching statistical methods by rules?
Dear Colleagues, I think it would help to draw a distinction between iron-clad rules and rules-of-thumb. Perhaps we can agree that it is generally not a good idea to teach students iron clad rules for making statistical decisions, especially if they do not understand the logic behind the rules. On the other hand, with experience we all develop rules of thumb that allow practical short cuts. If we can teach the logic behind the rules of thumb and the conditions under which the rules of thumb are likely to be valid and when they may fail, students can learn to use the rules of thumb effectively. It might be interesting to look at some of our favorite rules of thumb and see where they are likely to hold and where they are likely to fail (and how we can do diagnoses to tell the difference). How about this one: The sampling distribution of the mean is likely to be approximately normal with a sample of at least 30 cases IF the population is roughly symmetrical with no extreme outliers. Diagnosis: Plot the data, plus use whatever information is available about the population distribution. What do you think? Cheers, Dale Berger - Original Message - From: Robert Frick <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Sent: Thursday, December 16, 1999 8:14 PM Subject: Re: teaching statistical methods by rules? > I happened to have a vehement and probably radical opinion on this. > One of my sayings: "Ironically, our educational system is ideally suited > to teaching computers and ill-suited to teaching human beings." If you > are going to program a computer to do statistics, tell the computer > rules to follow. > > If you give students rules to memorize, they will surely forget them. > If you had a student who learned and applied the rules, people would say > that the student was mindlessly following rules and couldn't think for > him/herself. But your best student will just remember half the rules -- > and by that, I mean half of each rule. > > I know it is hard to make statistics fun, but FOLLOWING RULES IS NEVER > FUN. Not in math, not in games, nowhere. > > There are advantages to teaching rules. Most students like it. They > certainly understand that method of teaching. They just won't learn > anything. > > Bob F. > > > > EAKIN MARK E wrote: > > > > I just received a review which stated that statistics should not be > > taught > > by the use of rules. For example a rule might be: "if you wish to > > infer > > about the central tendency of a non-normal but continuous population > > using > > a small random sample, then use nonparametrics methods." > > > > I see why rules might not be appropriate in mathematical statistics > > classes where everything is developed by theory and proof. However I > > teach > > statistical methods classes to business students. > > > > It is my belief that if faculty do not give rules in methods classes, > > then > > students will infer the rules from the presentation. These > > student-developed rules may or may not be valid. > > > > I would be intested in reading what other faculty say about > > rule-based teaching depending on whether you teach theory or methods > > classes. > > > > Mark Eakin > > Associate Professor > > Information Systems and Management Sciences Department > > University of Texas at Arlington > > [EMAIL PROTECTED] or > > [EMAIL PROTECTED] >
interactive statistics tutorials and applets
Hi Robert, We have a website featuring interactive applets with tutorials for some introductory concepts. The applet for power is particularly cool - the user can control effect size, n, or power, and see dynamic connections. The URL is wise.cgu.edu These are free to use, though we appreciate being told about usage. Cheers, Dale Berger Dale Berger Professor and Chair, Psychology Claremont Graduate University 123 East Eighth Street Claremont, CA 91711 FAX: 909-621-8905 Phone: 909-621-8084 http://www.cgu.edu/faculty/bergerd.html - Original Message - From: Robert Lundquist <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Sent: Wednesday, December 15, 1999 5:33 AM Subject: WebCT in statistics courses? > Are there any statistics teachers who are constructing courses or course > components in WebCT? I would very much like to get in contact with you, > because there must be lots of material - quizzes, links,... - we could > share. > > By the way, does anyone know why I can't subscribe to edstat-l? No > response whatsoever when I try to get in contact with the listserv. > > --robert > > Robert Lundquist > Div of Quality Technology & Statitics > Lulea University of Technology > Sweden > > [EMAIL PROTECTED] >
Re: ANOVA with proportions
Just a reminder that transformations can be used on proportions as a dv to reduce the skew, important if some values approach 0 or 1. These include arcsine, probit, and logit. Each needs special treatment when p=0 or p=1. Cohen and Cohen (2nd ed. of Applied MR/C) has a section on transformations for proportions (pp. 265-270). Cheers, Dale Berger William B. Ware wrote: > As I recall, there was an article by Lunney et al that appeared in the > Journal of Educational Measurement that examined the use of ANOVA with "1" > and "0" as the DV. I believe that they concluded that distortion was > minimal when the distributions were within an 80/20 split... I think that > the article was in the early 70s, perhaps 1971. > > As Don has noted, proportions are means... which will be symmetrically > distributed when the split is about 50/50. Apparently, the Central Limit > Theorem applies as long as sample size is sufficiently large... > > Bill > > __ > William B. Ware, Professor and Chair Educational Psychology, > CB# 3500 Measurement, and Evaluation > University of North Carolina PHONE (919)-962-7848 > Chapel Hill, NC 27599-3500 FAX: (919)-962-1533 > http://www.unc.edu/~wbware/ EMAIL: [EMAIL PROTECTED] > __ > > On Tue, 14 Dec 1999, Robert Dawson wrote: > > > > > - Original Message - > > From: Donald F. Burrill <[EMAIL PROTECTED]> > > To: Wouter Duyck <[EMAIL PROTECTED]> > > Cc: <[EMAIL PROTECTED]> > > Sent: Tuesday, December 14, 1999 9:03 AM > > Subject: Re: ANOVA with proportions > > > > > > > On Tue, 14 Dec 1999, Wouter Duyck wrote: > > > > > > > I have a question. I have n subjects. For each subject, I have a > > > > proportion. I want to test if there are some differences in that > > > > proportion, depending on some independent variables (e.g. sex) on which > > > > the subjects differ. > > > > > > > > Can I use those proportions as a dependent variable in an ANOVA? > > > > > > Why not? Proportions are means, after all. Might even be more > > > interesting analyses to be pursued, if the proportions represent (or, > > > perhaps, conceal?) some repeated measures on the subjects. > > > > My first thought was that this seemed like a rather cavalier misuse of > > ANOVA, given that the population distributions are rather far from normal, > > and that Bernoulli distributions have a relation between mu and sigma that > > ANOVA fails to exploit. However, out of curiosity, I ran the following > > simulation twenty times: > > > > MTB > random 10 c11; > > SUBC> bernoulli 0.4. > > MTB > random 10 c10; > > SUBC> bernoulli 0.5. > > MTB > random 10 c12; > > SUBC> bernoulli 0.6. > > MTB > stack c10-c12 c13; > > SUBC> subs c14. > > MTB > oneway c13 c14 > > MTB > table c13 c14; > > SUBC> chisquare. > > > > and a similar one in which the null hypothesis was true 80 times, and > > discovered that the p-values obtained are actually rather close! The main > > peculiarity of the distribution of the ANOVA p (if Ho is true) is that it is > > very granular at the high end: the value 1.000 appeared several times, as > > did several other values. The chisquare test seemed to have slightly more > > power, but not by as much as I'd expected. > > > > I still think that chi-square is probably a better choice,and logistic > > regression more flexible - but I was surprised how well the screwdriver > > drove the nail... > > > > -Robert Dawson > > > > > >
Re: Student Ratings
A relevant article on biases in our judgments of what we learn is the following: Bjork, Robert A.(1995). Memory and metamemory considerations in the training of human beings. In J. M. A. P. Shimamura (Ed.), Metacognition: Knowing about knowing (pp. 185-205). Cambridge, Massachusetts: The MIT Press. Bjork has done several studies that show student ratings of how much they learned to be quite in error, especially with respect to knowledge retained after several weeks or more. Students feel they have learned more when the presentation is simple and easy to understand, and they feel they have learned less when they must struggle more with the material. (I'm sure that some types of confusion are more productive than others!) Cheers, Dale Berger Professor and Chair, Psychology Claremont Graduate University 123 East Eighth Street Claremont, CA 91711 FAX: 909-621-8905 Phone: 909-621-8084 http://www.cgu.edu/faculty/bergerd.html - Original Message - From: Ed Bailey <[EMAIL PROTECTED]> To: Ed-Stat List <[EMAIL PROTECTED]> Sent: Tuesday, November 30, 1999 9:06 AM Subject: Student Ratings > By way of introduction -- > > I teach Elementary Statistics at a Two-year college. (MS - > Statisics, PhD - Math Education). > > Question -- > On page 489 of "Elementary Statistics" by Triola (7th), there is a > sidebar article on " Student Ratings of Teachers". However, it does not > give any information on the studies used for the claim that student > evaluations have a negative correlation with amount of material learned. > If you happen to know the article refered to or any related studies, > please let me know. > Thank You -- Ed Bailey <[EMAIL PROTECTED]> > > >