## User Tag List

1. Originally Posted by BlueGray
How is your information cleaner? His information was created using actual data whereas yours has no basis.
WTF?

330 INTP Normal sample

Total INTP sample 911

581/911 = 64%

I used the actual data - worekd back from the % to give me people, he used the reported % which is is not as clean. Either way it's bollox anyways because we have questionable sample etc

Make sense...?

2. Originally Posted by tinkerbell
WTF?

330 INTP Normal sample

Total INTP sample 911

581/911 = 64%

I used the actual data - worekd back the % to give me people, he used the reported % which methematically is not as clean. Either way it's bollox anyways

Make sense...?
Those are two separate samples. You have 582/4829 of gifted are INTP and 330/9320 of people are INTP. If only 1% of the total population is gifted then 3.4% of INTPs are gifted. There was no gifted/total values given.

Check and 3.4/1 is the value in Avis's chart.

The main problem is you assume that the gifted sample and the normal sample are two parts of one sample which they are not. The normal sample is just a database sample of types while the gifted is a sample of people in a gifted group.

Edit: Also while working with counts is cleaner than percentages counts created from the percentages are no cleaner than the percentages themselves. The counts are only cleaner if they are the values the percentages are from and not vice-versa.

3. Originally Posted by tinkerbell
Big picture is that I would think even with screwy definitions, and all the other issues.. the only things that you could vaguely rely on NP thing (skew was big enough to indicate it's got a lot of wiggle room for failure) and the overall N thing (which was a massive skew)....
you wont like that... the findings are large enough differences that they would hold true with slightly better defintions
In your mind. You do realize that you're essentially saying "Despite all the screwy data, and the fact that the study is completely invalid, I still think the results will hold true in another study."...What makes you think that? The data and the study? Or your personal bias of what consistitutes "giftedness" and who is likely to have it?

4. Originally Posted by BlueGray
Those are two separate samples. You have 582/4829 of gifted are INTP and 330/9320 of people are INTP. If only 1% of the total population is gifted than 3.4% of INTPs are gifted. There was no gifted/total values given.

Check and 3.4/1 is the value in Avis's chart.

The main problem is you assume that the gifted sample and the normal sample are two parts of one sample which they are not. The normal sample is just a database sample of types while the gifted is a sample of people in a gifted group.
In terms of INTPs in terms of the joint population its a different statistic than the first list I produced, which was careful labled as beign the proportion of all INTPs who were gifted (64%), not the proportion of the whole sample who were INTPs.

I didn't assume gifted and norm as the same sample, I did exactly the opposite, if you look a the calculation above its pretty clear I didn't..

And what makes you think the two samples are seperate because there is nothing in the report that says so - which I've looked through but not at oober depth.. Which is part of the issue... I will look more fully tomorrow but it's late here.

Of the total population

Norm sample = 14148

= 4.1%... which was why I said AVIS stats were wrong.. because he divided %/% which is a bit shonky... gets you a ball park but not the absolute number

5. Originally Posted by Lauren Ashley
In your mind. You do realize that you're essentially saying "Despite all the screwy data, and the fact that the study is completely invalid, I still think the results will hold true in another study."...What makes you think that? The data and the study? Or your personal bias of what consistitutes "giftedness" and who is likely to have it?
yes, I work with quant data samples - there is a limit to how much a source will be screwed up..

In a normal population the data skews 70+% S type

In the gifted sample it's 70+% N type

So yea I'd say that aint gonna change an awful lot once the definition is tied down, P is less pronouced... but still fairly big...

No No no - I didn't say the study was invalid, I said the sample was badly defined/reported... It may be that once the defintion of gifted is established and the sample is established as independnat, it will be valid.

PS for what it's worth, at these type of sample sizes very small changes are signficant.

6. *sigh* Okay. You're just doing exactly what I said -- seeing what you want to see. There is a limit to how much data can be skewed, but you don't just throw out some parts of the study while keeping others. Unless you are biased.

7. Originally Posted by Lauren Ashley
*sigh* Okay. You're just doing exactly what I said -- seeing what you want to see. There is a limit to how much data can be skewed, but you don't just throw out some parts of the study while keeping others. Unless you are biased.
No you are not understanding... but thats OK, the survey isn't nessesarily invalid, just badly reported - to the point it is vague on some fundementals.

The really big differences are likely to still hold true... which is a pragmatic view in light on the lack of definitions.... given the scale of the differences.

An example... if I find 3 or 4 qualitative surveys where XYX appear more pronounced in a group... so long as those 3 or 4 surveys are well conducted, there is a fair to middling chance it wills tack up in quant research... there is no absolute gaurantee, but fairly often... now we have 14 surveys with 4000odd people that say N is more prevelent... I'd say that is likely to hold true... pragmatically speaking...

Maybe if you feel the need, you could do some primary research of your own?

8. Originally Posted by tinkerbell
In terms of INTPs in terms of the joint population its a different statistic than the first list I produced, which was careful labled as beign the proportion of all INTPs who were gifted (64%), not the proportion of the whole sample who were INTPs.

I didn't assume gifted and norm as the same sample, I did exactly the opposite, if you look a the calculation above its pretty clear I didn't..

And what makes you think the two samples are seperate because there is nothing in the report that says so - which I've looked through but not at oober depth.. Which is part of the issue... I will look more fully tomorrow but it's late here.
How can 64% of all INTPs be gifted if gifted are only 1% of all people? if all gifted were INTPs there would only be 28% of INTPs as gifted.

The norm group(n = 9,320) is composed of high school students in 11th-12th grades. Data for the norm group is adapted from the Atlas of Type Tables(Macdaid, Kainz & McCaulley, 1986).
Source of the normative values.

The literature review was done by means of the online version of the Educational Resource Information Center (ERIC) and Dissertation Abstracts International. Currently, ERIC contains either abstracts, full texts of studies, or both indexed from 1966 to the present. Keywords used in the search with various combinations were gifted, talented, personality, personality characteristics, personality types, psychological types, Myers-Briggs Type Indicator, and MBT I. Four hundred and twelve studies either in full-text or in abstract format were found. After an examination of each abstract, 63 studies were selected for further review. The rest of the studies were excluded from further investigation for three possible reasons: They were completely irrelevant to this research, they did not use the MBTI, or they were not original research.

After 63 studies we re obtained, including articles, reports, books, and dissertations, they were coded in identification forms for further review, which indicated that only 14 of them had enough data for inclusion. Each study had to report either the number of participants falling into each type, the eight basic personality types of the participants, or both to be included in this research synthesis. The 14 studies yielded 19 independent samples because some of them had more than one sample. Also, multiple studies by an author were carefully reviewed to avoid duplication in the synthesis. When sample characteristics matched in different studies by an author that were published in different journals and at different times, the one that had more data about findings and sample characteristics was included in the synthesis. Only two studies of one author (Mills, 1984; Mills & Parker, 1998) were included because there were 14 years between these two studies and the sample characteristics were significantly different. The 19 samples were then coded in sample characteristics forms and type distributions forms for inclusion.

There were two completely separate methods used to gather the normative information and the gifted information.

This finds the percentage of INTPs in either source that were gifted.

If I took a sample of 10 people with 2 INTPs and a sample of 100 gifted people with 10 INTPs your method would give that 83% of INTPs are gifted. You are treating sample size as an indicator of the prevalence of giftedness.

9. Originally Posted by BlueGray
How can 64% of all INTPs be gifted if gifted are only 1% of all people? if all gifted were INTPs there would only be 28% of INTPs as gifted.
Ok one step at a time...

How many people are INTP in the normative sample?

10. Originally Posted by tinkerbell
In terms of INTPs in terms of the joint population its a different statistic than the first list I produced, which was careful labled as beign the proportion of all INTPs who were gifted (64%), not the proportion of the whole sample who were INTPs.

I didn't assume gifted and norm as the same sample, I did exactly the opposite, if you look a the calculation above its pretty clear I didn't..

And what makes you think the two samples are seperate because there is nothing in the report that says so - which I've looked through but not at oober depth.. Which is part of the issue... I will look more fully tomorrow but it's late here.

Of the total population