From: email@example.com Subject: Re: Troll alert - parallel thread Date: 01 Jun 1999 00:00:00 GMT Message-ID: <firstname.lastname@example.org> References: <yEsU2.email@example.com> <firstname.lastname@example.org> <awFU2.23$Sc5.email@example.com> <firstname.lastname@example.org> <DSOX2.2563$Sc5.email@example.com> <firstname.lastname@example.org> <6S1Y2.2936$Sc5.email@example.com> <firstname.lastname@example.org> <VDVZ2.12$8N2.email@example.com> <firstname.lastname@example.org> <email@example.com> <firstname.lastname@example.org> <email@example.com> <firstname.lastname@example.org> <MW003.24$Z01.email@example.com> <firstname.lastname@example.org> <email@example.com> <firstname.lastname@example.org> <37440F20.DFB41E77@fmr.com> <email@example.com> <firstname.lastname@example.org> <email@example.com> <firstname.lastname@example.org> <email@example.com> <firstname.lastname@example.org> <email@example.com> <firstname.lastname@example.org> <email@example.com> X-Http-Proxy: 1.0 x39.deja.com:80 (Squid/1.1.22) for client 188.8.131.52 Organization: Deja.com - Share what you know. Learn what you don't. X-Article-Creation-Date: Tue Jun 01 16:18:07 1999 GMT Newsgroups: alt.education,alt.dads-rights.unmoderated,alt.parents.solutions X-Http-User-Agent: Mozilla/4.0 (compatible; MSIE 4.01; Windows 98)
In article <firstname.lastname@example.org>, email@example.com wrote: > In article <firstname.lastname@example.org>, > email@example.com wrote: > > In article <firstname.lastname@example.org>, > > email@example.com wrote: > > > In article <firstname.lastname@example.org>, > > > WStreett@shell.monmouth.com (Wilbur Streett) wrote: > > > > email@example.com wrote: > > > > >Frankly, judging from your previous assertion that "this data > > doesn't > > > > >correlate", the odds of your even understanding any of the above > > are > > > > >REAL small. Nonetheless, it will be shared with you if you so > > > request, > > > > >in order to give you the opportunity to prove that there are some > > > > >electrons wiggling around that cranium. > > > > > > > > > >The page http://fathersmanifesto.com/crosslinktimss1.htm > contains > > > some, > > > > >but not ALL, of the assumptions, some of which are assumptions > > about > > > why > > > > >the UN data differs so much from the TIMSS data. > > > > > > > > Naw, he won't answer that.. he didn't answer my explaination of > the > > > mental > > > > age being a factor in the test scores either. > > > > > > > > Wilbur > > > > > > Agreed, Wilbur. He never responds when a fact which disputes his > > > feminist perspective of the world is posted, so here is the specific > > > disconnect between the TIMSS data base and the UN Statistical > > Abstract, > > > 37th Edition http://fathersmanifesto.com/crosslinktimss1.htm > > > > > > The TIMSS data shows that the percent of men teachers is HIGHER than > > the > > > UN data shows it to be, by: > > > > > > Switzerland 20% > > > Japan 3% > > > Norway 30% > > > Sweden 36% > > > Spain 33% > > > Iran 3% > > > Canada 4% > > > Hong Kong 34% > > > New Zealand 2% > > > France 24% > > > Korea 3% > > > England 33% > > > Ireland 18% > > > Singapore 11% > > > Portugal 13% > > > > > > And the TIMSS data shows the percent of men teachers to be LOWER > than > > > the UN data by: > > > > > > Iceland 10% > > > Cyprus 6% > > > Belgium 13% > > > Czech Republic 48% > > > Hungary 4% > > > Russia 57% > > > > > > Some of these smaller differences could be because of the different > > > years the data is collected, different definitions, etc. But WHAT > > can > > > explain differences of 30% and 40% and even 57%??! > > > > > > > Thank you, John, for this excellent example of your ineptitude, > alluded > > to above. > > > > The reason for the difference is because they are two different > > populations. One is a measure of the percent of teachers of a country > > that are male. The other is a measure of the percent of teachers of a > > country WHOSE CLASSES TOOK THE TIMSS who are male. These are not the > > same group. > > > > > WHICH data is accurate? Why does the TIMSS data have such low > > > correlation between TIMSS scores and the percent of male teachers, > > > > BOTH data sets are accurate. YOU are confused. The TIMSS data shows > > 'such a low correlation' because there is no significant correlation > > between the scores and the gender of the teacher. > > > > Let me try to explain this using very simple concepts. Try to keep up, > > John. > > > > Let C represent the total teacher population of a country. > > Let T represent that percent of C that actually participated in the > > TIMSS study. T is a subset of C, therefore T does not equal C. In > fact, > > T will be significantly smaller than C. > > > > The UN data represents the percent of C that are male. > > The TIMSS data represents the percent of T that are male. > > > > These are INDEPENDENT measures. There is no reason to expect a > > relationship between the two. There is no indication that the TIMSS > > study attempted to select classes that reflected the overall > percentage > > of teachers in the country. Furthermore, C represents ALL of the > > teachers, where T represents the teachers of SPECIFIC GRADE LEVELS. > > Remember, there are actually 3 T samples (4th, 8th, 12th). Actually, > > there are more as there were multiple tests at the 12th grade level. > > > > One does not expect to find the same results for two independent > > samples. You might as well ask why the percentage of male teachers in > > Japan is different from the percent of male teachers in Korea. > > > > Since you are postulating that there is a relationship between TIMSS > > scores and the gender of the teacher, the ONLY data set that is > > applicable is the one that comes with the TIMSS data. > > > > This data HAS NOT BEEN PUBLISHED for the 12th grade tests. > > This data SHOWS A COEFFICIENT OF 0.06 for the 8th grade sample. > > I do not remember off the top of my head the relationship at the 4th > > grade. Seems to me that almost all of the countries had a majority of > > teachers as females. > > > > Is this elementary concept clear to you now, John? The only > > 'discrepancy' is in your ability to understand basic statistics. > > > > James Powell > > Whew, James, you had better stick to something simpler than teaching. > > You are 100% wrong, unless you now assert, AFTER you have made the above > blunder, that you originally intended to point out that those who take > the TIMSS are not randomly selected, or are specially selected for some > characteristic. > > If this subset is randomly selected, then the percent of male teachers > in the subset will be within 2-3% of the percent of male teachers in the > total population. In the case of TIMSS, there is no reason to merely > ASSUME that they were specially selected, which makes it a bit difficult > for you to now make that claim. >
You miss the obvious point above, nimrod. While the selection of the final students WAS random, the selection of the level to test was not. Your data from the UN is for ALL of the teachers in the country. The data from the TIMSS was for the GRADE LEVEL that participated in that test.
Let's illustrate this, shall we?
From the SASS (Schools and Staffing in the United States: A Statistic Profile, 1993-94)
In the Public school arena, the reported percentage of male teachers is 27.1% for this survey. This represents your 'UN statistic'.
The percentage of male teachers at the elementary level is 16.2% (averaged between the three number reported). Let this represent the 'Primary TIMSS statistic'.
The percentage of male teachers at the secondary level is 46.7% (averaged again). This can represent the 'TIMSS 8th grade statistic'.
I have just illustrated a variation of 10.9% to 19.6% from the National statistic. There are no published data (to my knowledge) that breaks down the UN statistic by educational level. Furthermore, the US example above only breaks by broad level - the results for an individual grade level could be much greater (or lesser) than illustrated above.
Do you now understand why you can not use the UN data to support your claim? You are trying to use an indirect statistic to prove a direct relationship. This is especially wrong in the case of the 4th and 8th grade data when the actual numbers are available.
> This part of TIMSS was essentially a poll, and polls like the Gallup > Poll can determine the characteristics of all 270 million Americans to > an accuracy within 3%, by polling only 1,000 households--a subset of > only .001% of the total population of households. >
As stated before: This was NOT a poll, it was a survey. The data was NOT used to extrapolate to a national number, unlike a Gallup Poll. It was not intended for that purpose and there is no evidence that the criteria for conducting such a poll were followed. It was not the purpose, so the data may not be used to serve that purpose.
> You say: > > > These are INDEPENDENT measures. There is no reason to expect a > > relationship between the two. > > This is a completely remarkable statement, Jimmy. Even IF the subset > were specially selected (i.e., if those who participated in TIMSS were > only the honor students of the country--which is not the case with > TIMSS), the two samples would be neither independent nor unrelated. You > could claim that you didn't know what the numeric relationship is, but > even that would be wrong, because you have enough data in TIMSS to > calculate that relationship. And even if that data weren't available, > there is no definition in the world, math or otherwise, which would > enable you to claim them to be "independent". When one is a subset of > the other, they are DEPENDENT by definition. >
Again: Look up the definition of dependent and independent in relation to mathematics and scientific variables.
> And IF there is a difference between the percent of men teachers who > teach the top scoring students within a country and the percent of men > teachers who teach the lower scoring teachers, then both of us know what > that relationship would be. Right, James? >
No, we don't. Produce the data and we can see what the realtionship is.
> > I do not remember off the top of my head the relationship at the 4th > > grade. Seems to me that almost all of the countries had a majority of > > teachers as females.
A quick check of the data reveals that only 2 of the 26 countries participating in the primary (4th grade) TIMSS reported over 50% of the teachers as male.
<random noise deleted>
Sent via Deja.com http://www.deja.com/ Share what you know. Learn what you don't.