This reply is based on the data from this link, http://cloud9.norc.uchicago.edu/faqs/nels.htm not the PDF. I hit 'reply' on the last message.
On Wed, Jul 21, 2010 at 12:16 PM, Scott Stroz <boyz...@gmail.com> wrote: > Sorry, maybe I am just dumb, but I do not see how that makes any > sense. Here is what does not compute with me (not saying its wrong, > just does not make sense) > > 1. None of the followups include the same people. For example, the > first follow up does not include data from parents. > 2. Beyond the first follow-up it seems they added new people to the study > 3. The 3rd follow up only includes about 15,000 people or so, while > the base year included about 34,000 (students, teachers, parents, etc) > (a decrease of about 56%). How accurate can the results be when the > last stage includes less than half of the original sample? > > This is why I hate statistics, its like Calvin-ball. You make up the > rules as you go along, usually to make the numbers say what you want > them to say. > > Lastly, you originally said this study was over 14 years with 50,000 > children and took place in Chicago? The link you provided indicates it > was about 25,000 students, over 6 years and does not really indicate > where they were from - except that some were from 'Vermont, > Connecticut, New York and Washington'. Maybe I missed something in > between? > > > On Wed, Jul 21, 2010 at 11:56 AM, Larry C. Lyons <larrycly...@gmail.com> > wrote: >> >> OK Scott last thing. >> >> This is the 4th follow up technical report on the methodology: >> http://nces.ed.gov/pubs95/95426.pdf >> >> On Wed, Jul 21, 2010 at 11:47 AM, Larry C. Lyons <larrycly...@gmail.com> >> wrote: >>> Also I found some more detail explanation of the followups for the >>> survey which ended in the late 80's. From what I understand the same >>> methodology was used. >>> http://cloud9.norc.uchicago.edu/faqs/nels.htm >>> >>> On Wed, Jul 21, 2010 at 11:40 AM, Larry C. Lyons <larrycly...@gmail.com> >>> wrote: >>>> Here Scott, this is a very brief explanation of longitudinal research: >>>> http://en.wikipedia.org/wiki/Longitudinal_study >>>> >>>> To get how the researchers did the actual assessments, you'll need to >>>> troll the NORC site's methodology sections for how they handled >>>> attrition etc. >>>> >>>> >>>> On Wed, Jul 21, 2010 at 8:31 AM, Scott Stroz <boyz...@gmail.com> wrote: >>>>> >>>>> Why do you seem to be taking that personally? >>>>> >>>>> I am just curious how following some kids for 1 year and some for 14 >>>>> years can yield consistent data. As I said, a lot can happen to people >>>>> in 14 years. In 14 years you can go from 2nd grade to college >>>>> graduate. Or from 6th grade to being a doctor. I understand that it >>>>> would be difficult to follow all the children for the same period of >>>>> time, but it just seems like a pretty wide disparity, especially with >>>>> children. A lot happens in 14 years with children. >>>>> >>>>> Maybe I am just an idiot, but I cannot seem to find anything at the >>>>> link you post even linking to the study you mentioned. >>>>> >>>>> On Wed, Jul 21, 2010 at 8:18 AM, Larry C. Lyons <larrycly...@gmail.com> >>>>> wrote: >>>>>> >>>>>> Its still legitimate. The longitudinal sampling techniques took such >>>>>> into account. Go to the site and look at how they do that sort of >>>>>> research. I'm pretty satisfied with their methodology, as is the >>>>>> entire field. You need to do your own research about it. I don't see >>>>>> why I ought to provide freebies when I charge a consulting fee for >>>>>> doing such. >>>>>> >>>>>> On Tue, Jul 20, 2010 at 5:04 PM, Scott Stroz <boyz...@gmail.com> wrote: >>>>>>> >>>>>>> 1-14 years? That seems to be a pretty big disparity for some kids >>>>>>> compared to another. A lot of shit (good and bad) can happen to a >>>>>>> person in 14 years. How can those numbers even be remotely accurate?. >>>>>>> >>>>>>> On Tue, Jul 20, 2010 at 4:10 PM, Larry C. Lyons <larrycly...@gmail.com> >>>>>>> wrote: >>>>>>>> >>>>>>>> No its simply a fact, not an excuse. For instance take the NORC >>>>>>>> dataset (see http://www.norc.org/homepage.htm) - this data is the >>>>>>>> result of a 20 year longitudinal study of all the children in the >>>>>>>> Chicago region school systems, including urban, suburban and rural >>>>>>>> systems. The children were followed throughout their school career. In >>>>>>>> the end over 50,000 children were followed for about 1 to 14 years. >>>>>>>> Not only was school achievement assess, but socioeconomic status, >>>>>>>> parental involvement, etc. >>>>>>>> >>>>>>>> The shared variance (or r squared value) between race and economic >>>>>>>> status was over 40%, meaning that the two factors (race and SES) were >>>>>>>> strongly related. To such an extent that you cannot statistically >>>>>>>> remove the effect of poverty from ethnicity effects nor can you >>>>>>>> eliminate the effects of race on effects due to socio-economic status. >>>>>>>> >>>>>>>> Similar results are found in the census data and in other very large >>>>>>>> datasets. Its not saying that one group is better than the other, its >>>>>>>> saying that this strong relationship exists and has to be taken into >>>>>>>> account in any statistical model you create. >>>>>>>> >>>>>>>> On Tue, Jul 20, 2010 at 3:31 PM, Jerry Barnes <critic...@gmail.com> >>>>>>>> wrote: >>>>>>>>> >>>>>>>>> "Race and poverty are real close. Real close. Really really close. So >>>>>>>>> close >>>>>>>>> together that its really really really difficult to remove the >>>>>>>>> effects of >>>>>>>>> one from the other." >>>>>>>>> >>>>>>>>> >>>>>>>>> One of the most racist ideas I have heard or read. It's that >>>>>>>>> sentiment that >>>>>>>>> gives people an excuse for failure. I can't succeed because my skin >>>>>>>>> color >>>>>>>>> is [fill in the blank]. >>>>>>>>> >>>>>>>>> >>>>>>>>> >>>>>>>>> J >>>>>>>>> >>>>>>>>> - >>>>>>>>> >>>>>>>>> No greater injury can be done to any youth than to let him feel that >>>>>>>>> because >>>>>>>>> he belongs to this or that race he will be advanced in life >>>>>>>>> regardless of >>>>>>>>> his own merits or efforts. - Booker T. Washington >>>>>>>>> >>>>>>>>> >>>>>>>>> >>>>>>>> >>>>>>>> >>>>>>> >>>>>>> >>>>>> >>>>>> >>>>> >>>>> >> >> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~| Order the Adobe Coldfusion Anthology now! http://www.amazon.com/Adobe-Coldfusion-Anthology-Michael-Dinowitz/dp/1430272155/?tag=houseoffusion Archive: http://www.houseoffusion.com/groups/cf-community/message.cfm/messageid:323349 Subscription: http://www.houseoffusion.com/groups/cf-community/subscribe.cfm Unsubscribe: http://www.houseoffusion.com/groups/cf-community/unsubscribe.cfm