The Fox News Version of Events

[Al Doggity]

No we are not.  You were arguing that ten people out of thousands were used to calculate the statistics.  This is false. 


http://www.getbig.com/boards/index.php?topic=561440.50


In this thread  you continue to argue that I'm wrong about the ten series incidents.

I'm not arguing anything. I'm trying to get your position nailed do because it keeps changing.

Are you saying that the study limits series incidents to 10 as not to skew the data?
Also, do you agree that there were 150,000 respondents to the survey?

Or are these wrong?

[Archer77]

You repeatedly confused samples cases with total number of respondents.

[Al Doggity]

I actually think you are right about series incidents. I just want to make sure I understand you correctly.

[240 is Back]

WTF are you talking about.  Who do you think won the 2 elections before Obama?
Are you seriously going to act like it hasn't been a flip flop for well over a century as to which party controls the Executive?
Seriously.  

I'm not sure you realize this... but...

Dems have won 4 of 6 elections for POTUS... by VERY comfy  margins.
Repubs have won 2 of 6 elections for POTUS... both times by 1 state that could have gone either way.  if FL is allowed to recount, Gore wins, history shows that now.  And OHIO, lol, Repubs SUED so a recount coudln't start, which may have explained the epic exit polls showing Kerry winning big there.

Without 2 shady court decisions, Dems are fucking 6 for 6 in POTUS elections.   Repubs can't be cocky for 2016.

[Al Doggity]

You repeatedly confused samples cases with total number of respondents.

I think I just didn't understand you.

I am trying to understand your position now.

Were there 150000 respondents to the survey?
Did the survey limit series incidents to 10 so they wouldn't skew the survey?

[Archer77]

I'm not arguing anything. I'm trying to get your position nailed do because it keeps changing.

Are you saying that the study limits series incidents to 10 as not to skew the data?
Also, do you agree that there were 150,000 respondents to the survey?

Or are these wrong?

It hasn't changed since December after I reread the material and then read more information.  I mentioned that I reread the material in December and return to my original argument.

You said 10 respondents or ten total people in the survey were plucked from the thousands not a cap of 10 on series incidents.  Are you reading anything I've posted in just the last ten minutes.

[Al Doggity]

It hasn't changed since December after I reread the material and then read more information.  I mentioned that I reread the material in December and return to my original argument.

You said 10 respondents or ten total people in the survey were plucked from the thousands not a cap of 10 on series incidents.  Are you reading anything I've posted in just the last ten minutes.

So, in other words  you agree that the survey included 150000 respondents and that survey also limited series incidents to 10 so that the data was not skewed. Good, so we're on the same page with this?

[Archer77]

so, then  do you agree or disagree with this statement:

150,000 people were interviewed for dozens of categories. Some categories have less than 10 respondents. The study warns that those categories are unreliable.


True or false?

Here you are again claiming ten respondents or ten people out of thousands. This is a total misinterpretation of the information.

I think I just didn't understand you.

I am trying to understand your position now.

Were there 150000 respondents to the survey?
Did the survey limit series incidents to 10 so they wouldn't skew the survey?

As I've told you a thousand times. As I told you in this very thread.  Its not ten out of one hundred and fifty thousand.

This is how it works, Kore.  Limiting the series incidents to ten is necessary to prevent national data from being skewed.  For example, a female in a household (wife or a child) is repeatedly raped by a father, brother or uncle-whatever.  In that situation the incidents of rape are numerous because they occur over years. That data is good for determining the statistics for that particular kind of rape.  However, it becomes problematic when that data is used to determine national statistics on the total incidents of rape for the year.  To prevent errors in the national data they limited the number of incidents in the above scenario to ten.   This was my original argument and the argument I came back to after I reread the material. I diverted briefly from this argument after giving you the benefit of the doubt.  When I reread the information I realized I was right with my earlier argument.

All the supporting info is in the other thread.  There is no 10 respondents, meaning only ten people responded.   The word isn't even used in the link you posted.  You misrepresented the information in the other thread by claiming the survey only included ten total respondents when in fact it had thousands.  

[Archer77]

so, then  do you agree or disagree with this statement:

150,000 people were interviewed for dozens of categories. Some categories have less than 10 respondents. The study warns that those categories are unreliable.


True or false?

Once again, your bullshit train hurdles forward. "Series Incidents"  means multiple incidents perpetrated on the same victim/s.

That paragraph has nothing to do with case studies of less than 10. It just means that if someone is beaten up by the same  person  or something like that, the report doesn't count it more than 10 times.
 

10 incidents in a series and 10 respondents isn't he same thing.

[Archer77]

Hulk got it

Al Doggity I think you may be incorrectly conflating the warning about N=10 on the frontpage with this specific method that also uses the number 10 as the limit to how many reports can come from an individual.  

If I'm understanding it correctly, if someone reported 1-10 incidents in the study period they would all get counted, if someone reported 11 or more, only the first 10 are counted.  So if anything it is under-estimating crime, and is not related to the statistical problems we pointed out earlier with N≤10 for these types of things in any obvious way that I can see.  It is just a way of avoiding having the results be rendered invalid or not generalizable, due to a small number of outliers that may report hundreds of incidents.

I've lost track of what's going on and who is fucking who and whose mother so I'll contribute to this thread in another fashion:

[Al Doggity]

This is how it works, Kore.  Limiting the series incidents to ten is necessary to prevent national data from being skewed.  For example, a female in a household (wife or a child) is repeatedly raped by a father, brother or uncle-whatever.  In that situation the incidents of rape are numerous because they occur over years.

Great, so we're both in agreement that series incidents are limited to 10 so that they don't  skew the data, correct?  I feel like you're avoiding saying yes or no for some reason  

[Skip8282]

I'm not sure you realize this... but...

Dems have won 4 of 6 elections for POTUS... by VERY comfy  margins.
Repubs have won 2 of 6 elections for POTUS... both times by 1 state that could have gone either way.  if FL is allowed to recount, Gore wins, history shows that now.  And OHIO, lol, Repubs SUED so a recount coudln't start, which may have explained the epic exit polls showing Kerry winning big there.

Without 2 shady court decisions, Dems are fucking 6 for 6 in POTUS elections.   Repubs can't be cocky for 2016.

Nonsense and much too narrow a view.  Over the past century or so, the Repubs have held it for about 60 years and same with Dems.  It's a typical cycle (one that needs to be broken).

Claiming either side can't win ignores even a basic historical analysis.

[Archer77]

Great, so we're both in agreement that series incidents are limited to 10 so that they don't  skew the data, correct?  I feel like you're avoiding saying yes or no for some reason  

Thats not what you argued.  You argued it was ten total respondents. You outright dismissed the argument that series incidents are limited to 10 so they don't skew national data. Below is the exchange where you roll your eyes at series incidents.

Yes, but they are referring to the incorporation of serial victimization and its effect on national victimization rates.


Now when the NCVS national victimization rates are estimated
to include series victimizations, the experiences of all series
victims will be taken into consideration. When series victims
state that the number of times the victimization occurred is
10 or fewer, those experiences will be counted at their stated
value using the victim’s response provided when first asked to
report this count. Series victims who provide responses that
are greater than 10 will have their experiences counted as 10
victimizations so that the overall impact on the victimization
rates of the higher and less consistent estimates will be reduced.
Series victims who are unable to provide a count of the
number of times the victimization occurred, but who report
that it occurred at least six times, will be counted as having
experienced 6 victimizations (the modal response category).

Okay. So, right from your link:

In 2013, series incidents accounted for about 1% of all victimizations and 4% of all violent victimizations.

So , statistically insignificant. Time to swerve again.

[Al Doggity]

Hulk got it

Exactly, so what Huk is saying is that the series incidents are limited to 10. I'm not sure why you won't just say, yes or no   Is that correct or not?

[Archer77]

Exactly, so what Huk is saying is that the series incidents are limited to 10. I'm not sure why you won't just say, yes or no   Is that correct or not?

See above post.  You dismissed the argument you are now claiming to support.  In the post from December, I argue that series incidents are limited to ten.  You claimed it's irrelevant.   Here is the exchange again

Yes, but they are referring to the incorporation of serial victimization and its effect on national victimization rates.


Now when the NCVS national victimization rates are estimated
to include series victimizations, the experiences of all series
victims will be taken into consideration. When series victims
state that the number of times the victimization occurred is
10 or fewer, those experiences will be counted at their stated
value using the victim’s response provided when first asked to
report this count. Series victims who provide responses that
are greater than 10 will have their experiences counted as 10
victimizations so that the overall impact on the victimization
rates of the higher and less consistent estimates will be reduced.
Series victims who are unable to provide a count of the
number of times the victimization occurred, but who report
that it occurred at least six times, will be counted as having
experienced 6 victimizations (the modal response category).

Okay. So, right from your link:

In 2013, series incidents accounted for about 1% of all victimizations and 4% of all violent victimizations.

So , statistically insignificant. Time to swerve again.

I already posted that.  This is all you need to know to understand what the ten sampling size is and why it is used.   Whats so hard about understanding this?  I didn't make it up.  This is from their website. I think you're playing dumb on purpose to derail the thread.  It's not ten people you moron.

In 2012, series incidents accounted for about 1% of all victimizations and 4% of all violent victimizations. Weighting series incidents as the number of incidents up to a maximum of 10 incidents produces more reliable estimates of crime levels, while the cap at 10 minimizes the effect of extreme outliers on the rates.

Victimization weights used in analysis of NCVS data account for the number of persons present during an incident and for high-frequency repeat victimizations (or series victimizations). Series victimizations are similar in type but occur with such frequency that a victim is unable to recall each individual event or describe each event in detail. Survey procedures allow NCVS interviewers to identify and classify these similar victimizations as series victimizations and to collect detailed information on only the most recent incident in the series.

The weight counts series incidents as the actual number of incidents reported by the victim, up to a maximum of 10 incidents. Including series victimizations in national rates results in rather large increases in the level of violent victimization; however, trends in violence are generally similar regardless of whether series victimizations are included.

The warning is to tell the reader that the series incidents might be greater than 10.

Fine. You know what series incidents are. This whole discussion stems from you posting info from this study and claiming that certain black-on-white crime was at epidemic levels while white-on-black crime in those categories was statistically zero. When I pointed out that these stats were derived from fewer than 10 cases, you claimed that didn't make a difference. Right in the link you provided, it says categories with fewer than 10 case studies are unreliable. It literally says that. It's simply common sense. More case studies makes a survey more reliable.


None of this has anything to do with series incident. You keep posting that blurb, but it is completely irrelevant. Limiting the number of series incidents to 10 has nothing to do with the reliability of 10 or fewer case studies.

Here you dismiss case incidents and again claim that the stats are wrong because only ten individual cases studies were used.  Meaning, only ten respondents not ten incidents in a series. It's time for you to concede

[Al Doggity]

Thats not what you argued.  You argued it was ten total respondents. You outright dismissed the argument that series incidents are limited to 10 so they don't skew national data. Below is the exchange where you roll your eyes at series incidents.

I feel like we might have misunderstood each other. I just want to make sure I understand your position. You're saying that SERIES INCIDENTS are limited to 10, correct?

And that the study surveyed more than 150,000 people?

Or are these statements wrong?

[Archer77]

I feel like we might have misunderstood each other. I just want to make sure I understand your position. You're saying that SERIES INCIDENTS are limited to 10, correct?

And that the study surveyed more than 150,000 people?

Or are these statements wrong?

Read the above exchange as well as the screencaps I posted.  You claimed ten total respondents were taken out of the thousands which is incorrect. You dismissed the argument that the ten was in reference to series incidents. Your words.

[Al Doggity]

See above post.  You dismissed the argument you are now claiming to support.  In the post from December, I argue that series incidents are limited to ten.  You claimed it's irrelevant.   Here is the exchange again

I guess I"m not going to get you to clarify your position. Gee, I wonder why 

Ok, so in December you claim that series incidents are limited to 10 and that I said the opposite? Hmm, why is it that in OCTOBER,  I had to explain what "series incidents" meant to you:

http://www.getbig.com/boards/index.php?topic=551654.msg7773110#msg7773110

Ok, now let's deal with the part where you seem to be having the most confusion. What they are talking ab out in the blurb that you don't understand is "series victimization". They even define it for you:

high-frequency repeat victimizations (or series victimizations). Series victimizations are similar in type but occur with such frequency that a victim is unable to recall each individual event or describe each event in detail.

Ok. So, now we know that "series victims"= people who are repeated victims of similar crimes. Keeping that in mind, the following passage that you didn't seem to understand should now make sense to you:

The weight counts series incidents as the actual number of incidents reported by the victim, up to a maximum of 10 incidents.

So, what that means is that the report counts up to 10 incidents in a series, even if there are more.

Let me know what about this you don't understand? I can be even clearer if necessary.

It's interesting that so many of the things that you claim are correct and that you said are things that I actually said. That's WEIRD!

[Archer77]

I guess I"m not going to get you to clarify your position. Gee, I wonder why 

Ok, so in December you claim that series incidents are limited to 10 and that I said the opposite? Hmm, why is it that in OCTOBER,  I had to explain what "series incidents" meant to you:

http://www.getbig.com/boards/index.php?topic=551654.msg7773110#msg7773110



It's interesting that so many of the things that you claim are correct and that you said are things that I actually said. That's WEIRD!

I already clarified my position several times.  You have a track record of not reading any information. The fact remains you thought the survey statistics were calculated by taking ten individuals or respondents from the 150,000.  This is untrue.  Respondents was a word you repeatedly used.  The word did not appear in the link you provided to prove the existence of the word.  The exchange with AJ clearly shows you truly did believe it was only ten respondents not ten incidents in a series.

You also dismissed the argument of ten incidents here.

Yes, but they are referring to the incorporation of serial victimization and its effect on national victimization rates.


Now when the NCVS national victimization rates are estimated
to include series victimizations, the experiences of all series
victims will be taken into consideration. When series victims
state that the number of times the victimization occurred is
10 or fewer, those experiences will be counted at their stated
value using the victim’s response provided when first asked to
report this count. Series victims who provide responses that
are greater than 10 will have their experiences counted as 10
victimizations so that the overall impact on the victimization
rates of the higher and less consistent estimates will be reduced.
Series victims who are unable to provide a count of the
number of times the victimization occurred, but who report
that it occurred at least six times, will be counted as having
experienced 6 victimizations (the modal response category).

Okay. So, right from your link:

In 2013, series incidents accounted for about 1% of all victimizations and 4% of all violent victimizations.

So , statistically insignificant. Time to swerve again.

[Al Doggity]

I already clarified my position several times.  You have a track record of not reading any information. The fact remains you thought the survey statistics were calculated by taking ten individuals or respondents from the 150,000.  This is untrue.  

It is untrue. I also never said it. You did. In the link you provided I was making fun of you.
http://www.getbig.com/boards/index.php?topic=551654.msg7773196#msg7773196

The report questioned a total of more than 10(150,000 or more) but uses ten as the sample size to correct statistical imbalances. Get it straight, the 10 are a random sampling of the larger pool of individuals surveyed.  

In the very next post I point out that that makes no sense.  In the conversation with aj. I later point out that this is nonsense you spewed.

[Archer77]

It is untrue. I also never said it. You did. In the link you provided I was making fun of you.
http://www.getbig.com/boards/index.php?topic=551654.msg7773196#msg7773196

In the very next post I point out that that makes no sense.  In the conversation with aj. I later point out that this is nonsense you spewed.

In the post from December, I argue that series incidents are limited to ten.  You claimed it's irrelevant.   Here is the exchange again

Yes, but they are referring to the incorporation of serial victimization and its effect on national victimization rates.


Now when the NCVS national victimization rates are estimated
to include series victimizations, the experiences of all series
victims will be taken into consideration. When series victims
state that the number of times the victimization occurred is
10 or fewer, those experiences will be counted at their stated
value using the victim’s response provided when first asked to
report this count. Series victims who provide responses that
are greater than 10 will have their experiences counted as 10
victimizations so that the overall impact on the victimization
rates of the higher and less consistent estimates will be reduced.
Series victims who are unable to provide a count of the
number of times the victimization occurred, but who report
that it occurred at least six times, will be counted as having
experienced 6 victimizations (the modal response category).

Okay. So, right from your link:

In 2013, series incidents accounted for about 1% of all victimizations and 4% of all violent victimizations.

So , statistically insignificant. Time to swerve again.

I already posted that.  This is all you need to know to understand what the ten sampling size is and why it is used.   Whats so hard about understanding this?  I didn't make it up.  This is from their website. I think you're playing dumb on purpose to derail the thread.  It's not ten people you moron.

In 2012, series incidents accounted for about 1% of all victimizations and 4% of all violent victimizations. Weighting series incidents as the number of incidents up to a maximum of 10 incidents produces more reliable estimates of crime levels, while the cap at 10 minimizes the effect of extreme outliers on the rates.

Victimization weights used in analysis of NCVS data account for the number of persons present during an incident and for high-frequency repeat victimizations (or series victimizations). Series victimizations are similar in type but occur with such frequency that a victim is unable to recall each individual event or describe each event in detail. Survey procedures allow NCVS interviewers to identify and classify these similar victimizations as series victimizations and to collect detailed information on only the most recent incident in the series.

The weight counts series incidents as the actual number of incidents reported by the victim, up to a maximum of 10 incidents. Including series victimizations in national rates results in rather large increases in the level of violent victimization; however, trends in violence are generally similar regardless of whether series victimizations are included.

The warning is to tell the reader that the series incidents might be greater than 10.

Fine. You know what series incidents are. This whole discussion stems from you posting info from this study and claiming that certain black-on-white crime was at epidemic levels while white-on-black crime in those categories was statistically zero. When I pointed out that these stats were derived from fewer than 10 cases, you claimed that didn't make a difference. Right in the link you provided, it says categories with fewer than 10 case studies are unreliable. It literally says that. It's simply common sense. More case studies makes a survey more reliable.


None of this has anything to do with series incident. You keep posting that blurb, but it is completely irrelevant. Limiting the number of series incidents to 10 has nothing to do with the reliability of 10 or fewer case studies.

Here you dismiss case incidents and again claim that the stats are wrong because only ten individual cases studies were used.  Meaning, only ten respondents not ten incidents in a series. It's time for you to concede

[Al Doggity]

In the post from December, I argue that series incidents are limited to ten.  You claimed it's irrelevant.   Here is the exchange again


Here you dismiss case incidents and again claim that the stats are wrong because only ten individual cases studies were used.  Meaning, only ten respondents not ten incidents in a series. It's time for you to concede

I dismissed SERIES INCIDENTS because they weren't related to the discussion. This was the point of contention:


http://www.bjs.gov/index.cfm?ty=dcdetail&iid=245#Methodology

In cases where the CV is greater than 50%, or the unweighted sample had 10 or fewer cases, the estimate is noted with a “!” symbol (Interpret data with caution. Estimate based on 10 or fewer sample cases, or the coefficient of variation is greater than 50%).

You claimed that samples with fewer than 10 cases were completely, infallibly accurate even though the people who conducted the study warned that they are extremely unreliable. You only started posting that nonsense about SERIES INCIDENTS because the number 10 was convenient for you.

[Archer77]

Depends, is that sample group composed of Sheldon Adelson and 9 other of America's richest people?

Another way of looking at it is by asking if the 538 presidential electors that make up the electoral college are a reliable-enough group to chose our president every 4 years (Keeping in mind that there are 316 Million Americans)?

No is the straightforward answer to your question. A sample of 10 people can not give you a reliable national statistic.

"1"

Exactly. But more importantly... that's exactly what quoted text from the study is saying, too.

You confirm you misinterpreted the number ten to refer to ten people not ten incidents in a series. Then you claim the study itself says its limited to ten people.

[Archer77]

I dismissed SERIES INCIDENTS because they weren't related to the discussion. This was the point of contention:


http://www.bjs.gov/index.cfm?ty=dcdetail&iid=245#Methodology
You claimed that samples with fewer than 10 cases were completely, infallibly accurate even though the people who conducted the study warned that they are extremely unreliable. You only started posting that nonsense about SERIES INCIDENTS because the number 10 was convenient for you.

As I mentioned before I corrected myself after I gave your argument of ten meaning ten people not incidents the benefit of the doubt.  Series incidents are important to the study because as I said many times they help eliminate outliers that skew the national data.  After I corrected myself you continued to argue that it was ten respondents not ten incidents in a series. You believed the study was inaccurate because the sample size was to small.  When I read further I discovered that the sample size was pretty large and the number ten was in reference to series incidents. The below thread is the important one, in particular the first page were you discuss sample size and respondents with aj and OMR.

http://www.getbig.com/boards/index.php?topic=561440.0

[Archer77]

I'm not confusing series incidents with weighing methodology.  The link I posted discusses this.  Series incidents were effecting total counts.  This was my original argument.  

http://www.ncbi.nlm.nih.gov/books/NBK202273/

No, it wasn't. You just desparately glommed onto this blurb because it contained the number 10. It has nothing to do with the margin of error, which is pretty straightforward.

Again you deny series incidents are an issue.  I admitted and corrected my errors back in December.  Even in this thread you repeated the mistake of 10 respondents.